Merge pull request #1 in PROCESSOR-SDK/pdm-anomaly-detection from pdm-demo-bringup to master

* commit '7439abc642c7a7faf8c9262ebdc26a311a82027c':
  Add ONNX model and python scripts for converting it to LSTM_model.h
  Add PdM anomaly detection code

19 files changed, 1802 insertions, 0 deletions

diff --git a/LICENSE b/LICENSE
new file mode 100644
index 0000000..f6cb9de
--- /dev/null
+++ b/LICENSE
@@ -0,0 +1,19 @@
+Copyright (c) 2016 Robert W. Rose
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/LSTM_model.h b/LSTM_model.h
new file mode 100644
index 0000000..2d42757
--- /dev/null
+++ b/LSTM_model.h
@@ -0,0 +1,77 @@
+/*This is the header file generated from the ONNX model*/
+
+#include <string>
+#include <vector>
+
+unsigned int NUM_LAYERS = 2;
+unsigned int DENSE_WEIGHTS_ROWS = 31;
+unsigned int DENSE_WEIGHTS_COLS = 2;
+unsigned int DENSE_BIASES_SHAPE = 2;
+unsigned int DENSE_ACTIVATION = 1;
+
+std::vector<float> DENSE_WEIGHTS = {
+0.210532277822, -0.846990406513, 0.55079549551, 0.0386005304754, 0.460019826889, -0.389443337917, -0.422828495502, -0.658000648022, -0.411930412054, -0.0299967080355, 0.0646565034986, -0.127355322242, -0.136991396546, -0.19303496182, 0.537825167179, -0.346626967192, -0.474057734013, 0.453149110079, 0.228727906942, 0.493378400803, 0.214479982853, 0.577478289604, 0.506602048874, 0.0788790583611, 0.271695792675, -0.262315154076, -0.65589261055, 0.1702619344, -0.455700933933, 0.282108694315, 0.377390474081, 0.149042129517, -0.33547013998, 0.464447796345, 0.19582682848, 0.535690009594, 0.368645846844, -0.560067415237, -0.268278330564, 0.503697216511, -0.222755014896, 0.405264765024, 0.240972399712, -0.269863516092, 0.299161523581, 0.103564433753, 0.0878492891788, -0.223425522447, -0.245210498571, 0.0221647024155, -0.532086789608, -0.0356784388423, 0.34632819891, -0.265819728374, -0.215275481343, -0.961182057858, -0.611859202385, -0.143221437931, 0.05197237432, -0.196741729975, 0.404271095991, -0.517124414444};
+
+std::vector<float> DENSE_BIASES = {
+-0.279018938541, 0.107939668};
+
+std::vector<unsigned int> LSTM_RECURRENT_ACTIVATION = {4};
+std::vector<unsigned int> LSTM_ACTIVATION = {4};
+std::vector<unsigned int> RETURN_SEQUENCES = {0};
+std::vector<unsigned int> N_HIDDEN_NEURONS = {31};
+std::vector<unsigned int> W_i_ROWS = {2};
+std::vector<unsigned int> W_i_COLS = {31};
+std::vector<unsigned int> W_f_ROWS = {2};
+std::vector<unsigned int> W_f_COLS = {31};
+std::vector<unsigned int> W_c_ROWS = {2};
+std::vector<unsigned int> W_c_COLS = {31};
+std::vector<unsigned int> W_o_ROWS = {2};
+std::vector<unsigned int> W_o_COLS = {31};
+std::vector<unsigned int> U_i_ROWS = {31};
+std::vector<unsigned int> U_i_COLS = {31};
+std::vector<unsigned int> U_f_ROWS = {31};
+std::vector<unsigned int> U_f_COLS = {31};
+std::vector<unsigned int> U_c_ROWS = {31};
+std::vector<unsigned int> U_c_COLS = {31};
+std::vector<unsigned int> U_o_ROWS = {31};
+std::vector<unsigned int> U_o_COLS = {31};
+std::vector<unsigned int> b_i_SHAPE = {31};
+std::vector<unsigned int> b_f_SHAPE = {31};
+std::vector<unsigned int> b_c_SHAPE = {31};
+std::vector<unsigned int> b_o_SHAPE = {31};
+std::vector< std::vector<float> > W_i = {
+{0.185322135687, -0.824516355991, -0.642393827438, -0.425822228193, 0.293285936117, -0.0121912341565, -0.324504047632, 0.597941279411, 0.311987578869, -0.136105984449, -0.0732097104192, 0.770671188831, -0.205151304603, 0.176611080766, 0.0811026617885, 0.845845758915, 0.771258175373, 0.199620261788, -0.151814773679, 0.265885561705, 0.608966231346, 1.08486700058, -1.02740645409, -0.0338779725134, 0.0174377560616, -0.676254630089, 0.171720176935, -0.744176924229, -0.0273252110928, 0.208006575704, 0.110072128475, -0.401547402143, -0.515101671219, -0.118443764746, -0.337929040194, 0.159724146128, -0.050756521523, 0.907452821732, -0.162338942289, -0.555124700069, -0.620603978634, -0.0859199911356, -0.0510654933751, -0.0102122807875, -0.30154183507, 0.0126685164869, -0.079158090055, -0.00292449118569, 0.237848713994, 0.0574446730316, -0.263453274965, 0.441671043634, 0.200348317623, 0.187376976013, -0.218234792352, 0.103939890862, -0.317922919989, 0.378643244505, 0.329482734203, -0.473105251789, -0.825138568878, 0.0711314529181}
+};
+std::vector< std::vector<float> > W_f = {
+{0.332748025656, 0.0523226633668, 0.110590398312, 0.526660919189, 0.791073739529, 0.815366089344, 0.456697463989, 0.704801619053, 0.796210944653, 0.0887657403946, 1.12468755245, 0.268723905087, -0.327608138323, -0.0266997776926, 0.481298834085, 0.605446517467, 0.723925292492, 1.04179382324, -0.0929877460003, 0.153678372502, 0.653793334961, 0.59051245451, 0.0277767088264, -0.231102436781, -0.232954755425, -0.25073248148, 0.420234024525, 0.370259165764, -0.139870449901, -0.390542477369, 0.44000634551, -0.620040655136, -0.268983364105, 0.658930122852, 0.209272459149, -0.0116781275719, 0.718103706837, 0.523001134396, -0.616794168949, -0.12540872395, -0.816522777081, 0.23254366219, 0.433318346739, 0.840964078903, -0.397520244122, -0.773413360119, -0.227990701795, -0.188090503216, 0.0812743529677, -0.217632204294, -0.710024833679, -0.229106917977, 0.0362048633397, 0.205409690738, 0.39506110549, 0.904763877392, -0.0510291755199, -0.236016601324, 0.0760101228952, -0.151532202959, -0.861478626728, 0.700033843517}
+};
+std::vector< std::vector<float> > W_c = {
+{0.0956307649612, 0.613787412643, 1.42434096336, -0.465489715338, -0.885539829731, 0.331036388874, -0.463429629803, 0.559589505196, -0.934589862823, 0.663712322712, 0.298835992813, 1.22075688839, 0.483471244574, -0.574919819832, -0.423118859529, 0.714640617371, -0.599209070206, 0.239795416594, 0.3150434196, -0.357135206461, -0.994406998158, 0.951134383678, 0.22748029232, 0.156485080719, -0.0671564117074, -0.834566652775, 0.784602880478, -0.389289140701, -0.446090251207, 0.0358697026968, 0.446910351515, -0.699675858021, 0.750261664391, 0.343975454569, -0.624528586864, -0.0853450447321, -0.350698500872, -0.797592699528, -0.309429317713, 0.849446773529, 1.02368044853, 0.415024578571, 0.439539432526, -0.143217951059, -0.0407542400062, 0.360431134701, 0.38262706995, 0.5365011096, 0.70000731945, -0.644960641861, 0.839984714985, 0.877264797688, -0.265248954296, 0.658727347851, -0.200940534472, 0.212469920516, 0.0511242486537, -0.259432256222, -1.01890397072, -0.219125330448, -0.0610960572958, -0.687466144562}
+};
+std::vector< std::vector<float> > W_o = {
+{-0.107104726136, -0.0896070674062, 0.802965939045, 0.352801054716, 0.350219219923, -0.168809011579, 0.648496985435, 0.746658265591, -0.248712971807, -0.24700962007, 0.370593160391, 0.572165369987, -0.42929688096, -0.421578377485, 1.03107631207, 1.03904819489, 0.523328661919, 1.42379152775, 0.151075184345, -0.186686187983, 0.324157714844, -0.154081508517, -0.0407108552754, -0.32866358757, -0.201205551624, -0.466150105, -0.0200563259423, 0.190965130925, -1.13164412975, 0.0166868362576, -0.308687716722, -0.514899194241, -0.0124524720013, -0.122631326318, 0.785848140717, 0.213736400008, 0.17377141118, 0.0467304028571, 0.0383704453707, 0.731132745743, -1.19046068192, 1.179885149, -0.22209353745, 0.0215659141541, -0.905560016632, -0.0975127071142, -0.0893698260188, -0.245528966188, 1.36968410015, -0.545720994473, -0.177812322974, -0.600349009037, 0.52797806263, 0.148664355278, 0.844377994537, -0.188374742866, -0.306566596031, 0.277793824673, 0.39385086298, -0.360828429461, -1.31102252007, 1.04591810703}
+};
+std::vector< std::vector<float> > U_i = {
+{-0.164487153292, -0.407542705536, 0.159912630916, 0.648254752159, -0.613477647305, -0.0732728615403, -0.274545997381, -0.260433137417, 0.280366152525, 0.474827498198, 0.280240088701, 0.395428329706, -0.113590732217, -1.12173962593, -0.329467684031, -0.350164711475, 0.160477310419, 0.403092175722, -0.402937591076, 0.00753971841186, 0.394124239683, -0.5355052948, -0.237472698092, 0.467440456152, -0.510356009007, 0.242923140526, 0.142857924104, 0.19709187746, -0.109128497541, -0.993200421333, 0.140478551388, -0.35166734457, 0.0562370121479, -0.639412999153, -0.167009696364, -0.106686957181, -0.126806974411, -0.571713209152, 0.450123429298, -0.622551620007, -0.28208220005, -0.0408553704619, 0.302280783653, 0.480019360781, 0.260600477457, 0.245027571917, -0.231700494885, 0.292674720287, -0.446479588747, -0.0386877022684, -0.029840502888, 0.241011783481, 0.704230546951, -0.28801497817, -0.31289306283, 0.0571827180684, -0.372752815485, 0.246893778443, 0.0210091192275, -0.0565862804651, 0.347900778055, -0.900236427784, -0.242373347282, 0.134584665298, -0.263310015202, 0.182442843914, -0.0910202264786, 0.304554104805, -0.795127093792, -0.428028792143, 0.146787345409, 0.192136183381, 0.0959995016456, 0.0966757833958, 0.200424581766, -0.113789923489, -1.16422116756, -0.24228258431, -0.256587892771, -0.274459004402, -0.803272068501, 0.0119419684634, 1.09562063217, 0.0718587115407, -1.26335823536, 0.733490884304, 0.265326201916, 0.0797169804573, -0.299673199654, -0.466012209654, 0.634292542934, 0.597472548485, 0.463436990976, -0.0922856032848, 1.16615080833, 1.0626142025, 0.1977789253, -0.00269439537078, -0.285869479179, -0.0248880460858, 0.286503612995, 0.135255530477, -0.621304273605, -0.206383794546, 0.179388433695, -0.30302387476, -0.240797460079, 0.717433273792, -0.262774467468, 0.187226936221, -0.422373026609, 0.0253499988467, -0.212347686291, -0.357483774424, 0.0767034813762, 0.543270349503, -0.218572050333, -0.48947134614, 0.343642055988, 0.767906725407, 0.130830034614, -0.134268254042, -0.549683153629, 0.548970878124, 0.624590337276, 0.480495393276, 0.445949167013, 0.522306740284, 0.00411103619263, 0.248325422406, 0.00594625947997, -0.475243955851, 0.2216501683, -0.0844806209207, 0.388139039278, -0.572655200958, -0.666153311729, -0.0269224829972, -0.415426880121, -0.203630030155, -0.0548370666802, 0.207273036242, 0.630436182022, 0.588027536869, 0.088676713407, -0.766153216362, 0.040290530771, 0.57573735714, -0.107092857361, 0.441415220499, -0.410136133432, 0.891629874706, -0.327789664268, 0.0766619220376, 0.265039891005, 0.929680109024, 1.94677746296, 0.775543987751, 1.72443604469, -0.148989260197, -0.132808059454, -0.564696550369, -0.528271377087, 0.00811330694705, 0.682681143284, 1.41143941879, -0.357699811459, 0.0670791938901, 0.297949612141, -0.261132359505, 0.302658081055, 1.06131291389, 0.546736717224, 0.40866419673, 0.106745764613, -0.608365654945, 0.563972592354, 0.6762560606, 0.814812242985, 0.812520265579, 0.149787783623, 0.602243185043, 0.0675226300955, 0.473846673965, 0.0268353205174, 0.722383499146, 1.390671134, 0.655241549015, 0.281917184591, 1.06690776348, -0.767260313034, -0.0510017499328, -0.585513174534, 0.138773754239, -0.312110543251, 0.418485760689, 0.860759079456, -0.55674302578, 0.204360798001, -0.804131507874, 0.584833920002, 0.0693883597851, 1.15516602993, 1.04773974419, 0.760290145874, 0.888959050179, 0.149687379599, 0.47254306078, -0.0609300397336, 0.445295840502, -0.311582297087, -0.19494214654, 0.386239141226, -0.165093749762, -0.165460571647, -0.56527364254, -0.374315232038, -0.138946682215, -0.800044417381, -0.114970937371, -0.0521705858409, 0.305533736944, -0.066392108798, 0.000597951002419, -0.0874828025699, -0.154430449009, 0.115694373846, -0.150818437338, 0.383218079805, -0.511324822903, -0.344857662916, -0.433309078217, 0.263184338808, -0.121228545904, 0.170186474919, -0.236393168569, 0.21826466918, 0.756543636322, 0.264278173447, -0.633802652359, -0.147351667285, -0.207492679358, -0.0609193146229, 0.00532165169716, -0.355129539967, 0.390977740288, -0.58093214035, 0.45332044363, 0.479072213173, 0.630649805069, -0.0380194149911, 0.274149060249, 0.373511403799, 0.049011003226, -0.37916713953, 0.164433777332, 0.279291093349, 1.19503068924, 0.0742072761059, -0.430000394583, -0.21801880002, 0.851600646973, 0.524434328079, 0.265210330486, 0.419434785843, -0.722075998783, 0.582435309887, 0.147055476904, -0.319798052311, 0.434719651937, 1.38389933109, -0.20397682488, -0.823963522911, 0.411769390106, 0.465590059757, 0.960044682026, -0.51882314682, 0.939898014069, -0.163087695837, 0.485007911921, 0.29119476676, -0.48755633831, -1.22392225266, 0.120499685407, -0.00624776491895, -0.153248459101, 0.451457649469, -1.25283944607, 1.03268873692, -0.569376468658, 0.616575062275, -0.339915305376, 0.600001573563, 0.690176486969, 0.353133827448, 0.882952451706, -1.48962819576, 0.670821905136, 0.224207192659, 0.514059305191, 0.843888282776, -0.15627284348, -0.395652532578, 0.539095163345, 0.696803152561, -0.298561513424, -0.172457501292, -0.0899025052786, 0.390438437462, -0.114717833698, 0.294596195221, -0.45960226655, 0.141172647476, 0.750059485435, 0.0719481036067, 0.420206606388, 0.222784742713, -0.44137391448, 0.244588196278, 0.670756518841, 0.0992781817913, 0.262624591589, 0.707923710346, 0.147759243846, -0.749383866787, 0.882630169392, -0.397331625223, 0.808132231236, -0.251220315695, 0.449114501476, -0.111033506691, -0.631867706776, -0.601980805397, 0.190706953406, 0.671102404594, -0.0186868663877, -0.576015949249, 0.114028081298, 0.398586988449, 0.18180270493, 0.119044058025, -0.0966572165489, -0.711808145046, -0.0299291517586, 0.061357807368, 0.212572187185, 0.0517120175064, 0.000841927016154, -0.0779775232077, 0.0772511437535, 0.164951711893, -0.0630072057247, 0.350497245789, -0.222510799766, -0.309259176254, -0.45270383358, 0.11665610224, 0.114547811449, -0.0482596047223, -0.294239163399, -0.336119443178, 0.807297766209, 0.725023388863, -0.471796900034, 0.182807207108, 0.0225853677839, -0.425974756479, 0.431024640799, -0.654430389404, 0.629156529903, 0.0387883260846, 0.513399243355, 0.43688544631, 0.397288948298, 0.169491678476, 0.288903951645, 0.167713269591, -0.821794748306, -0.414580583572, 0.14986024797, -0.476276963949, -1.17544054985, 0.107744380832, -0.326127499342, -0.514605522156, -0.0699354335666, 0.118005059659, -0.376910716295, 0.188896372914, 0.61270070076, 0.620187163353, 1.04264950752, 0.54366350174, 0.333110958338, -0.30417072773, 0.15594497323, 0.744870364666, -0.912702798843, 0.300396263599, 0.588330388069, 0.855659484863, -0.558107972145, 0.6105286479, -0.0117312464863, -0.434222072363, 0.0880842134356, -0.026462899521, 0.164516434073, 0.523595333099, 0.262407064438, -0.00614183349535, -0.504784703255, -0.220052748919, -0.1333951056, 0.901484549046, -0.409719914198, -0.1779910326, -0.0240657329559, 0.642626345158, -1.00084853172, 0.0544602125883, -0.0153825329617, -0.217744216323, 0.320210903883, -0.850863873959, 0.887492418289, -0.444515943527, -0.19537115097, -0.0819171667099, -0.747580111027, 0.150522857904, -0.0809876769781, 0.306356102228, -0.817374527454, -0.0831496715546, 0.683626651764, -0.104358166456, -0.215784758329, -0.364929646254, 0.0581333190203, 0.22409068048, 0.0140627184883, 0.0120426444337, 0.0403446182609, -0.0839538946748, -0.274827152491, 0.432591974735, 0.267429351807, 0.301961421967, -0.52459281683, 0.112877570093, -0.395714133978, 0.157957434654, -0.202919214964, -0.447013407946, -0.347543030977, 0.541160345078, 0.147864580154, 0.571336686611, -0.0288989376277, -0.277266472578, 0.596128940582, -0.244217306376, 0.818109214306, -0.374582409859, -0.503960847855, -1.12998652458, -0.298067510128, -0.374991089106, -0.327114880085, -0.364645600319, 0.0865003690124, 0.492246687412, 0.408382177353, 0.201397910714, -0.50353628397, 0.284086823463, 0.283284932375, -0.226132124662, 0.0656722337008, -0.43092417717, 0.0277040228248, -0.536834657192, -0.732004404068, -0.416810959578, 0.0975241810083, 0.936955571175, -0.794023096561, -0.172677233815, -0.0384489707649, 0.0177555419505, 0.440599888563, -0.268689066172, 0.542522311211, 0.183660060167, 0.397090375423, -0.0910913571715, 0.402351111174, -0.0908357128501, -0.174148067832, 0.0244021248072, 0.151753664017, 0.400920420885, 0.149080857635, -0.300013065338, -0.475381612778, -0.028100065887, -0.356042742729, 0.484075576067, 0.0740654245019, 0.378352046013, -0.315879017115, -0.126948535442, -0.12522906065, 0.175269782543, -0.0752810761333, -0.350499600172, -0.233585223556, 0.538995742798, -0.0347191430628, 0.443047851324, -0.300464123487, -0.337638616562, 0.556284606457, -0.256481736898, 0.12988114357, -0.478183090687, -0.0753780826926, -0.708748221397, 0.171034634113, -0.420573532581, 0.315279096365, 0.63263964653, 0.450077265501, 0.0206140894443, 0.286801755428, -0.249289557338, -0.525831580162, 0.517132937908, -0.126471325755, -0.0833186134696, -0.382867485285, -0.0735524222255, -0.494721859694, -0.324033051729, -0.0452652797103, 0.210131719708, 0.514053940773, -0.0110263107345, -0.956570386887, -0.339703917503, 0.208228453994, 0.0264142621309, 0.077341362834, 0.539237558842, 0.758221745491, -0.311438530684, 0.522510647774, -0.0517337098718, -0.708975791931, 0.435691028833, 0.287285894156, -0.549819648266, 0.0489376075566, -0.345976322889, -0.138600617647, 0.417212039232, 0.0213532391936, 0.0622787289321, 0.146891266108, -0.364064216614, -0.74403744936, -0.208205148578, 0.345515042543, 0.210370749235, 0.387875795364, -0.249161601067, 0.447263747454, 0.590879023075, -0.102956019342, -0.30993527174, 0.0979671478271, -0.462234795094, 0.0526035167277, 0.156474024057, -0.0577186346054, 0.100531190634, -1.19445288181, 0.126113474369, -0.0557793565094, 0.00289810402319, -0.300669878721, -0.523862719536, 0.547021925449, 0.0875920131803, 0.652453899384, 0.0691835731268, -0.207231938839, -0.771884024143, -0.309908807278, -0.777642190456, 0.246078327298, 0.211340993643, 0.189431622624, -0.0524844750762, -0.545048654079, -0.11482847482, 0.407315611839, 0.376557826996, -0.339429050684, -0.259641855955, 0.145675361156, -0.515954613686, 0.707189023495, -0.302177757025, -0.665942549706, 1.4158103466, -0.0196849759668, 0.380451321602, -0.056689646095, -0.51342433691, 0.102460436523, 0.533276855946, -0.661154747009, -0.1155404374, 0.214141726494, 0.788060307503, 0.309116214514, -0.347334384918, -0.757768213749, -0.285526365042, -0.321270138025, 0.517450809479, -0.279202580452, 0.507998287678, -0.475096255541, -0.33223977685, -0.228368580341, 0.0197465848178, -0.192769706249, -0.52666670084, 0.205842882395, 0.49493008852, -0.163632556796, 0.452393203974, -0.0329611040652, -0.181181535125, 0.181062892079, -0.305417060852, -0.485239565372, -0.819802105427, -0.336250424385, -0.404516607523, 0.0112056937069, -0.548414349556, 0.384314745665, -0.00844592228532, -0.176367342472, -0.166681304574, -0.536572158337, -0.455058038235, -0.127014622092, 0.749053537846, 0.00158268236555, -0.0975999906659, -0.601346969604, 0.30337318778, -0.220618084073, 0.153511270881, -0.400322914124, -0.494821369648, 0.589502334595, 0.793783426285, -0.167749494314, 0.358410328627, -0.0771593675017, -0.605319261551, 0.556215465069, -2.08747410774, 0.342077553272, 0.353174626827, 0.333294332027, -0.172520130873, -0.165387749672, -0.213544026017, -0.953555047512, 0.127406224608, 0.466204047203, -0.660779416561, 0.0761788710952, -0.0531936846673, 0.23602540791, -0.139079079032, 0.510132908821, -0.927884936333, 0.477215826511, 0.15670542419, -0.464307159185, 0.405231356621, 0.924576461315, -0.126855760813, 0.886300623417, 0.22820314765, 0.424698889256, 0.637558281422, -0.660719752312, 0.414879769087, 1.09232091904, -0.379344254732, 0.893301188946, -0.122210688889, 0.0453483648598, 0.824729979038, 0.0207926835865, 0.0403792373836, -0.0680488795042, -0.0480768457055, -0.319194704294, -0.112287148833, -0.369646340609, -0.0192338526249, -0.987628757954, -0.446307122707, 0.424960017204, -0.0494486391544, -0.270858675241, -0.414820849895, 0.309547901154, 0.0295022111386, 0.307085543871, 0.778900742531, -0.107606820762, -0.00586908869445, 0.113031215966, 0.0291796531528, 0.344350993633, 0.291163057089, -0.458524852991, -0.918518900871, -0.194868296385, -0.809211075306, -0.352026432753, 0.212357789278, -0.0104480003938, -0.877216100693, -0.286692887545, -0.24143640697, -0.496465653181, 0.228827640414, 0.100610949099, 0.233085379004, 0.314258515835, 0.41614151001, 0.376080274582, -0.478030204773, 0.410880923271, 0.12531080842, 0.42834675312, 0.0817243382335, -0.00883969664574, -0.881691634655, -0.570351064205, -0.637523770332, -0.495629072189, 0.291542351246, -0.435092538595, 0.221673086286, -0.366768181324, -0.160927116871, 0.67035150528, -0.615176916122, 0.267816901207, -0.190832167864, 0.447455942631, -0.224486753345, 0.675238370895, -0.683622658253, 0.247054189444, -0.0335351787508, -0.588586568832, 0.0193692930043, 0.294937759638, -0.0841782838106, -0.418257504702, -0.141904041171, 0.112735651433, -0.302489161491, -0.822585761547, 1.12332069874, 0.810719013214, 0.92630815506, 0.553161740303, 0.669064104557, 0.789175450802, 0.0595295540988, 0.328040480614, 0.142672494054, 1.26204276085, 0.703191518784, -0.319422125816, 1.00696074963, 1.56500065327, -0.180914118886, -0.27977013588, 0.212531447411, 0.922577500343, -0.0477373450994, -0.57782959938, 0.234480381012, -0.133011013269, 0.0124455792829, -0.0202074460685, -0.55104881525, 0.0753744319081, -0.0395135879517, -0.0474609397352, 0.0854560360312, 0.34555748105, -0.0113620730117, 0.108251556754, -0.0206192489713, 0.359737843275, 0.314718127251, 0.271229833364, -0.110280051827, -0.0929397493601, 0.277293533087, 1.07226550579, -0.282710403204, 0.23233217001, -0.168234527111, 0.109317548573, 0.968984484673, -1.5862557888, -0.162628695369, 0.0277825389057, -0.00271279597655, 0.632942199707, 0.546580553055, 0.143725767732, 0.41021296382, 1.04510653019, -0.323831290007, -0.496476531029, 0.293700605631, -0.27276724577, -0.304866164923, 0.223264828324, 0.315074622631, 0.106523774564, -1.04497349262, 0.640197455883, 0.410274893045, 0.6852222085, 0.322529107332, 0.651565909386, 0.739228010178, -0.047073662281, 0.272879123688, -0.150955498219, -0.236635446548, -0.168399587274, -0.268863230944, 0.255556851625, 0.035828627646, -0.179071620107, -0.343835622072, -0.0665528029203, -0.660620927811, 0.320234775543, -0.219391807914, -0.10191975534, 0.609591126442, 0.615102827549, 0.109470307827, -0.396675527096, 0.757627129555, -0.241236165166, -0.298540443182, -0.0442315675318, 0.313224673271, -0.024803051725, -0.839801967144, 0.129486396909, -1.34405696392, -0.416842848063, -0.511329770088, -0.320518761873, -0.162213683128, -0.56982165575, 0.0866421014071, -0.52089881897, 0.283288031816, 0.181297704577, -0.816463708878, 0.0414585545659, -0.168896883726, -0.131179049611, 0.252290189266, -0.354806959629, -0.0773894041777, 0.0907651484013, 0.803178548813, 0.123843699694, 0.17028132081, 0.226506382227, 0.314169317484, 0.175540968776, 0.0923963412642, -0.142761036754, -0.418033361435, 0.46586728096, -0.668575704098, 0.804650187492, 0.354451835155, -0.69633024931, 0.120617717505, 0.146716445684, -0.213579460979, -0.324909329414, -0.652023494244, -0.339829713106, -0.529457092285, -0.392566502094, -0.458715587854, -0.674575984478, 0.505500972271, 0.0594071224332, -0.105159506202, 0.347018748522, -0.413006663322, 0.828626155853, -1.09643268585, 0.0396186038852, -0.216222405434, -0.242173328996, -0.71623903513, 0.313321501017, 0.522072970867, 0.0715973451734, 0.330850750208, -0.662272572517, 0.379842311144, -0.00743951508775, 0.268450319767, -0.211613088846, 0.645617008209, -0.629408299923, -0.618042111397, -0.136103719473, -0.667649209499, 0.303324460983, -0.500855207443, 0.0376387983561, 0.198482051492, -0.310838550329, -0.689716219902, -0.496901482344, -0.100248679519, -0.291292279959, -0.95583409071}
+};
+std::vector< std::vector<float> > U_f = {
+{-0.579729378223, -0.240697517991, -0.391773641109, -0.0240022800863, -0.411112308502, 0.415624171495, -0.0047526913695, -0.269419491291, -0.368260592222, -0.812921464443, -0.109384700656, 0.0500135831535, -0.320259720087, -0.797238886356, -0.054978325963, -0.50250607729, -0.292500197887, 0.369918167591, -0.613622844219, -0.0181589126587, -0.188018590212, -0.582842886448, 0.0219135191292, 0.501310229301, -0.625574946404, -0.0332837402821, -0.290030449629, -0.0438743419945, -0.423029005527, -0.434717297554, -0.152828618884, 0.0307667832822, 0.187059700489, 0.143562138081, 0.467153996229, -0.200320243835, 0.324830114841, -0.187931135297, 0.0407067835331, -0.28892570734, 0.291180491447, 0.522511601448, -0.107481434941, -0.166256844997, 0.145582005382, -0.0777871981263, -0.579501390457, 0.0576013736427, 0.331325441599, 0.0442015863955, -0.237542748451, 0.125098735094, 0.146750941873, -0.0388164408505, 0.382148653269, 0.678474605083, 0.0561255514622, -0.249962866306, 0.49927714467, 0.490476608276, 0.18875233829, -0.152198985219, -0.189808696508, 0.272002488375, 0.223152041435, 0.114646427333, -0.45322740078, 0.219066694379, -0.386645436287, -0.423029690981, -0.12091396004, 0.0288026742637, 0.44174939394, -0.483458280563, -0.134082078934, -0.0279557500035, -0.704384088516, 0.12814360857, -0.254986315966, 0.433679521084, -0.570041418076, -0.0974168106914, 0.0302363745868, -0.433956056833, -0.789570391178, 0.99357765913, 0.738486230373, 0.162031829357, -0.832058727741, -0.414377868176, 0.439675629139, 0.107937663794, -0.343572825193, 0.0125817749649, -0.135692805052, -0.605703413486, 0.0514758676291, -0.210153132677, -0.572701394558, 0.617621719837, 0.279402196407, 0.286285996437, -0.08839494735, -0.436336070299, 0.266761064529, -0.137363597751, -0.324332386255, 0.431760847569, 0.911524832249, 0.288351804018, -0.608948409557, 0.184108152986, -0.0395081788301, 0.148599147797, 0.32482033968, 0.0214601289481, -0.499597609043, -0.504007339478, 0.531947016716, 1.00183415413, 0.244060441852, -0.234654963017, -0.284466505051, 0.92964053154, 0.146770775318, 0.366336911917, -0.0249889716506, -0.0122079923749, 0.217332869768, -0.110673725605, -0.117202416062, -0.241526171565, -0.167616009712, -0.891405284405, -0.31890463829, -0.27149784565, -0.09813477844, 0.264856249094, -0.346463918686, 0.659840703011, -0.228288516402, 0.235374420881, 0.370430916548, 0.167681753635, -0.342733711004, -0.387056231499, -0.925467073917, 0.344034463167, -0.341795682907, -0.108376920223, -0.709555804729, -0.311818391085, -0.0762023329735, -0.202133536339, 0.330305397511, -0.325266867876, 0.607773900032, -0.106952250004, 0.499734938145, -0.474227011204, -0.0895734354854, -0.141600623727, -0.501813828945, -0.298272758722, -0.13137434423, 0.610173463821, -0.0960054770112, -0.049923799932, -0.688569188118, -0.156922519207, -0.0889091566205, 0.176287263632, 2.88600850105, -0.314901798964, -0.573954343796, -0.768349409103, -0.299445867538, 0.0890558734536, 1.5026512146, -0.0780149847269, -0.628274023533, -0.641145825386, 1.57606589794, 0.726400136948, -0.664666235447, 0.276382386684, 0.551085412502, 0.290566623211, -0.366790264845, 1.15117657185, -0.444707900286, 0.0951429829001, -0.15641708672, 0.224048420787, -0.368727147579, -0.476319015026, 0.865709364414, -0.137042358518, -0.214546844363, -0.624815285206, 0.45021507144, 0.0785279050469, 0.0835598781705, 0.760127067566, 0.420478850603, 0.392497837543, -0.619300127029, -0.26378428936, -0.157186418772, 0.485694676638, -0.452785879374, 0.114374749362, -0.252173870802, 0.461731642485, -0.0714897140861, -0.13897703588, 0.188868910074, -0.128385365009, -0.657074272633, -0.243392631412, -0.0710246786475, -0.517055511475, 0.385283738375, 0.120848491788, -0.204280465841, -0.522168755531, -0.393322139978, 0.3502202034, -0.123526722193, -0.0982557758689, -0.40042924881, -0.229914024472, -0.494513064623, -0.321847587824, 0.194977909327, -0.397868782282, 0.067718654871, 0.131063088775, -0.0288040582091, -0.206563830376, 0.0182079095393, 0.271297633648, -0.179054841399, -0.264207214117, -0.69286519289, -0.418661177158, -0.262594968081, -0.568449556828, 0.269163787365, 0.137929677963, 0.0548856109381, 0.0584674961865, 0.41508808732, -0.875878870487, -0.586689352989, 0.0709109678864, 0.155825406313, 0.772737443447, -0.559619009495, -0.284445643425, 0.350556284189, 0.622111678123, 0.135816544294, 0.142503634095, 0.306301683187, -0.650290310383, 0.507817208767, -0.0417640991509, -0.234705746174, 0.188104748726, 0.905066311359, -0.239322468638, -0.0117977950722, 0.218682900071, 0.182186245918, 0.518645942211, 0.220537647605, 0.645013213158, -0.0675756558776, 0.46430426836, 0.206398323178, -0.562358379364, 0.313486456871, -0.798752069473, 0.489626199007, 0.0342953205109, 0.341822475195, -0.158351302147, 0.649500489235, 0.517660617828, 0.709735572338, 0.191699832678, 0.343735873699, 0.181281149387, 0.00268616294488, 0.755567312241, -0.478244155645, 0.617243826389, 0.253449887037, 0.963568270206, 0.397872716188, 0.273138165474, -0.261880934238, 0.60852009058, 0.535265922546, -0.69650042057, 0.858114480972, 0.744390308857, 0.233539000154, -0.229464411736, -0.647045850754, -0.0832387804985, 0.340829253197, -0.303880572319, -0.315676033497, 0.689193665981, -0.477797985077, -0.479350298643, -0.674902379513, -0.333230465651, -0.167654871941, 0.136966109276, 0.417031228542, -0.495346069336, -0.478774696589, -0.265086233616, -0.389864444733, 0.221837043762, -0.619465172291, -0.182093203068, -0.0861630067229, -0.60737067461, -0.784800291061, 0.170872405171, -0.111186757684, -0.870047688484, -0.866347312927, -0.551609218121, 0.242425218225, -0.242621302605, -0.725639164448, -0.246908560395, -0.376547545195, -0.680213928223, 0.184919148684, -0.408092558384, 0.476654082537, 0.21163316071, -0.109749421477, -0.247667014599, 0.102044299245, 0.65992307663, -0.322650372982, -0.164773181081, -0.798241734505, -0.296254038811, -0.382144272327, -0.165147379041, 0.603541791439, -0.427700132132, -0.123156808317, 0.190040752292, 0.300995379686, 0.232947990298, 0.483257859945, 0.230173125863, -0.676247477531, -0.159243002534, -0.414438933134, -0.0480983704329, -0.42405128479, -0.259621202946, -0.144406691194, 0.103743590415, 0.0430429130793, 0.616546213627, -1.23698651791, 0.558896720409, 0.268420279026, 0.0616095997393, -0.369732618332, -0.841801345348, 0.357456415892, 0.167683184147, 0.152035757899, -0.607366681099, -0.16161262989, -0.464338213205, -0.0268882736564, 1.47340667248, -0.36528813839, 0.36539876461, 0.395767390728, -0.0035567691084, -0.250310361385, 0.684129118919, 0.0685274973512, -0.180814534426, 0.00899717584252, 0.200045615435, 0.227090999484, -0.757442772388, -0.328870713711, 0.609381973743, -0.686259329319, 0.0379131250083, -0.251714229584, 0.341968119144, -0.419956296682, -0.183063656092, 0.0513952076435, -0.3560090065, -0.0566507615149, 0.0211090072989, 0.289371162653, -0.717945158482, 0.112896434963, 0.141625061631, 0.0602362863719, -0.308017522097, -0.646728754044, 0.609350085258, 0.544661879539, 0.252454012632, -0.200253546238, -0.491183161736, -0.376882940531, 0.649139404297, 0.412702530622, -0.128171145916, -0.255522102118, -0.344611078501, 0.669388175011, 0.0921176150441, 0.0969089344144, 0.655250191689, 0.415021896362, -0.0976783111691, 0.406340152025, -0.468279898167, -0.307071685791, 0.0665305256844, 0.287617266178, 0.81674182415, -0.390051871538, -0.0387920103967, -0.0731227695942, 0.473023176193, 0.0558998435736, 0.412459939718, 0.31103181839, -0.675665318966, 0.376408517361, 0.0329382121563, 0.0427423231304, -0.0929640829563, 0.225189611316, -0.027886737138, 0.124357715249, 0.236565813422, 0.304815322161, 0.615557909012, 0.448634266853, 0.454819411039, 0.389647275209, -0.033180937171, -0.701718509197, 0.096556827426, -0.131517618895, -0.42744666338, 0.196032688022, 0.0638964399695, 0.245661273599, 0.371226906776, -0.259907871485, 0.255794912577, -0.319425225258, 0.19839258492, -0.248832300305, 0.212510317564, -0.646145403385, 0.0419677756727, -0.183626249433, -0.51517701149, -0.117576546967, -0.198912248015, 0.526692569256, -0.00695317704231, -0.0801942497492, 0.41096329689, -0.135086968541, 0.569498598576, -0.423445791006, -0.0831989720464, -0.647928774357, -0.0476304143667, 0.211213365197, 0.574036061764, 0.0974889919162, 0.0040074181743, 0.340808212757, -0.28969591856, 0.22261941433, 0.194737628102, 0.209574952722, 0.441866368055, -0.274214893579, 0.220092102885, -0.0807638987899, 0.426905781031, 0.145819187164, 0.471180081367, 0.263562709093, -0.59473413229, 0.377283275127, 0.0465606302023, 0.00822502747178, 0.181930109859, 0.148309499025, -0.352126419544, 0.00746462587267, 0.301979899406, 0.2656917274, 0.715158879757, 0.304514884949, 0.382809013128, 0.293740808964, -0.661802053452, -0.617960214615, 0.401422053576, -0.617265880108, -0.369054377079, 0.369053930044, 0.309769362211, -0.353882044554, -0.140615493059, -1.03740394115, -0.355007350445, 0.352682679892, 0.40452259779, -0.223840564489, -0.439689874649, -0.567243874073, -0.149962797761, 0.0844186022878, -0.474122345448, 0.0611325576901, -0.0570296421647, -0.307366907597, -0.683964073658, 0.069020435214, 0.423344314098, -0.329758435488, -0.371612131596, -0.480321705341, 0.461175382137, -0.333911120892, -0.543587982655, -0.578899264336, -0.50201433897, -0.210972383618, -0.216104641557, -0.594712853432, 0.520093858242, 0.152158617973, -0.201858177781, -0.367576986551, -0.908140420914, 0.211481839418, -0.173844143748, -0.615172982216, -0.708517968655, -0.0951965451241, -0.273550033569, -0.275654107332, 0.512732207775, -0.566919386387, -0.0318128205836, 0.00217428128235, -0.211867466569, 0.129690721631, 0.269317597151, -0.417693078518, 0.00703907804564, -0.180487573147, -0.123906612396, -0.551772534847, -0.271103143692, -0.354702323675, 0.128252461553, -0.0385474413633, 0.106414414942, -0.247589573264, 0.00698315678164, -0.0887186080217, 0.294528365135, 0.180485606194, -0.274272561073, -0.0821488201618, -0.241977900267, 0.163563087583, 0.0889354273677, 0.394556999207, -0.0161370523274, 0.320505410433, 0.169003814459, -0.376825511456, 0.541992604733, 0.193778082728, 0.026314239949, 0.208189100027, 0.187958478928, -0.820899426937, 0.118026301265, -0.10349791497, 0.0394578352571, 0.0707554370165, -0.113599389791, 0.672230064869, -0.310120284557, 0.0898140221834, 0.350852161646, -0.203631088138, -0.123907469213, 0.203928560019, -0.251981019974, 0.281952947378, 0.243859305978, 0.121276192367, 0.445794135332, -0.314056575298, 0.0961595624685, -0.225795194507, 0.377101629972, 0.215090364218, 0.366714507341, 0.246535301208, -0.572864532471, 0.288375556469, 0.322296857834, 0.0266691055149, 0.31304949522, 0.042199190706, -0.455860853195, 0.334904879332, 0.232869058847, 0.287036687136, 0.312692373991, 0.0659090206027, 0.193059056997, 0.0881818458438, -0.0480946004391, -0.355062365532, -0.550945341587, 0.0731179788709, -0.354915201664, 0.27765712142, -0.0132603757083, -0.225266650319, -0.102469801903, -0.275519639254, 0.731525957584, -0.0246590003371, -0.299534469843, -0.225920438766, -0.228455320001, -0.327942460775, -0.278077065945, 0.54693543911, -0.194143459201, 0.0255733206868, -0.0270867347717, 0.503746747971, -0.0369442403316, 0.338712483644, 0.240481674671, -0.230860322714, 0.111184045672, -0.559044420719, 0.130134448409, -0.0424872450531, 0.00220347498544, 0.384383231401, -0.0701170191169, 0.622472584248, 0.185327991843, 0.360326111317, 0.0586520321667, 0.153322279453, -0.0570598803461, -0.504354476929, 0.692774772644, 0.39702039957, -0.155441090465, -0.456518620253, 0.034941226244, 0.0588169060647, -0.464926213026, 0.184596061707, 1.0513368845, -0.152978569269, -0.436923295259, 0.473860144615, 0.216794729233, 0.108939379454, 0.441177546978, 0.902628719807, 1.05245101452, -0.126265078783, 0.370298057795, -0.227412566543, -0.185615718365, 0.660686016083, 0.71403336525, 0.264653533697, -0.373310565948, 0.617018282413, 0.489990532398, -0.0935362055898, 0.0462726131082, 0.295728296041, -0.101450890303, 0.474373489618, 0.573818743229, -0.161887004972, -0.754527449608, 0.290287226439, 0.313316971064, 0.656889140606, 0.0972219333053, -0.0716303512454, 0.17260055244, 0.39352735877, 0.431635290384, 0.538031101227, 0.561284661293, -0.204435154796, -0.410260230303, -0.145853236318, 0.189406499267, 0.517140686512, -0.222079068422, 0.385745227337, 0.421404510736, 0.0715472251177, 0.10265763104, 0.419802725315, -0.473630845547, 0.254060775042, -0.632694244385, -0.208805173635, 0.0177669860423, -0.19705748558, 0.589730679989, -0.506344497204, 0.047658007592, -0.192752793431, 0.566913366318, 0.0692817866802, -0.246583238244, 0.0162540003657, -1.49162375927, 0.0491413548589, 0.150106698275, 0.399000257254, -0.152639985085, -0.246186479926, -0.59215515852, 0.437586903572, 0.373394668102, 0.305503159761, -0.61251950264, -0.241916432977, 0.390638768673, -0.624098658562, 0.0438898727298, -0.863707661629, -0.577601790428, -0.199276790023, 0.147946894169, -0.134792327881, -0.333698660135, 0.120271444321, 0.655990421772, -1.20608663559, -0.68566775322, 0.325098931789, -0.576812565327, -0.121651135385, 0.700207591057, -0.358716547489, 0.138019502163, 0.803461194038, 0.423798233271, 0.518534779549, -0.067420206964, -0.262941598892, 0.724140644073, 0.497597664595, -0.865565896034, -0.569445073605, 0.153895184398, 0.880490005016, -0.277198791504, 0.420099645853, -0.102995172143, 0.0871484652162, -0.53413105011, -0.212427526712, 0.486138731241, -0.174877434969, 0.0187892206013, 0.067981377244, 0.0417774654925, 0.182296946645, -0.592497587204, 0.556780397892, -0.175521165133, -0.264682203531, -0.0488861650229, 0.195661947131, -0.346799075603, -0.114110477269, 0.907555878162, -0.165068402886, -0.271177262068, -0.13010032475, 0.414531469345, -0.159666135907, 0.279220640659, 0.27878883481, -0.0521079413593, -0.00908483099192, 0.0546137131751, -0.221106931567, -0.259886592627, 0.247681424022, 0.36287266016, 0.245052456856, -0.810599148273, 0.745921313763, 0.208042383194, 0.0279369708151, 0.16183103621, 0.416752308607, -0.0265147443861, -0.268208086491, 0.49791431427, 0.455762803555, 0.0142332334071, -0.675378441811, 0.533760666847, 0.515494406223, 0.0783475190401, 0.433628559113, 0.558027625084, 0.366834670305, 0.354204088449, 0.115862496197, -0.0876316800714, -0.0248375348747, -0.504761517048, 0.48023557663, -0.141204029322, 0.663755238056, -0.0274682343006, 0.00642060115933, 0.0727510601282, -0.545299351215, -0.186217948794, -0.270892113447, -0.464565783739, -0.351237475872, -0.0882039666176, -0.181781664491, -0.564282953739, -0.573221623898, -0.320227593184, 0.250055462122, -0.332153171301, 0.000218620203668, -0.615575492382, -0.478811681271, -0.490893781185, -0.599205315113, -0.0468643195927, -0.37717962265, -0.465682476759, -0.476660132408, -0.803482353687, -0.232321605086, 0.0462054684758, 0.394775122404, 0.036762021482, -0.798236966133, -0.428599894047, -0.69172424078, -0.0666895136237, -0.539053976536, -1.49004662037, -0.280903965235, 0.149694919586, -0.526975214481, -1.11851418018, 0.148219779134, -0.267227470875, -0.404032796621, -1.52894961834, -0.54094350338, 0.0644419714808, -0.582399904728, -0.287099629641, -0.365315794945, 0.120949052274, -0.704033732414, -0.568006157875, -0.237801223993, -0.463440209627, -0.458468914032, -0.431300491095, -0.613944768906, -0.951498150826, -0.190325140953, -0.517063379288, -0.795393884182, -0.47432744503, -0.0943014621735, -0.278003245592, -0.0407003760338, -0.69560354948, -0.0227578505874, 1.11235868931, -0.207517802715, 0.463001102209, 0.103660002351, 0.060673866421, 0.38691174984, 0.234788775444, 0.143564611673, 1.44329452515, 1.06302022934, -0.469094455242, -0.132172107697, -0.0948769599199, 0.0846410617232, 0.385747849941, 0.351442813873, 0.0698994845152, -0.222074016929, -0.469291031361, -0.265184521675, 0.0497683249414, 0.35973444581, 0.30969157815, 0.111434161663, 0.523951888084, 0.092773668468, 0.280131906271, 0.386053591967, -0.415138632059, 0.108068540692}
+};
+std::vector< std::vector<float> > U_c = {
+{-0.476260900497, 0.363233804703, -0.0833832472563, -0.485053002834, -0.342705756426, 0.644412219524, -0.141952365637, -0.330390781164, -0.183823272586, 0.380058825016, 0.291497766972, -0.297042936087, 0.352544337511, -0.400815606117, 0.219869077206, 0.67021638155, -0.0782335475087, 0.460703641176, -0.603317856789, -0.0727846696973, -0.293882846832, -0.00808594282717, 0.192287102342, -0.636183381081, -0.295784294605, 0.18799290061, 0.219009041786, -0.14994302392, -0.113345332444, -0.545362830162, -0.296479672194, -0.0767128616571, -0.418924570084, 0.171866297722, 0.280353754759, 0.61053955555, -0.365708380938, 0.203083172441, -0.00563837448135, 0.399824887514, -0.256104558706, -0.0600364357233, -0.296908169985, -0.295463413, 0.260764271021, 0.0995672866702, -0.838992238045, -0.0102459685877, -0.140429675579, -0.00307480944321, 0.130755335093, -0.090179041028, 0.0287466980517, 0.244353726506, 0.0511544980109, 0.119147732854, 0.175830990076, 0.0809949710965, 0.665032565594, 0.599435567856, 0.526889681816, -0.318579733372, -0.088525891304, -0.378526329994, 0.480689913034, -0.302897900343, 0.873670756817, 0.439698904753, 0.141936421394, -0.236772105098, -0.336209595203, -0.2377910465, 0.0137182036415, -0.22522161901, 0.206687599421, 0.0298010352999, 0.244848906994, -0.174718618393, -0.0519334189594, 0.0680439993739, -0.232676193118, -0.304009646177, -0.549442529678, -0.301269084215, -0.221348389983, -0.684548377991, 0.0316169448197, -0.217143908143, -0.569811403751, 0.806551873684, 0.469743132591, -0.0124996053055, 0.0423225536942, -0.125226944685, 0.224116638303, -0.976225554943, -0.137934699655, -0.348645567894, -0.188023731112, -0.389280080795, 0.151754468679, -0.46952727437, -0.835674881935, 0.143766060472, 0.145547404885, -0.198087438941, -0.719173371792, -0.101115211844, 0.813134789467, -0.0540882870555, 0.0419979989529, -0.0675243362784, 0.357155710459, -0.0141962738708, 0.144741997123, -0.666016221046, 0.0315781757236, 0.417047381401, -0.37798666954, 0.382874697447, 0.713350594044, -0.256401181221, 0.182566106319, 0.215668484569, 0.62120449543, -0.0181763861328, -0.451997071505, 0.285670787096, 0.00353050790727, -0.14182458818, -0.699721038342, 0.0563862249255, -0.165772095323, 0.555769383907, -0.0757899656892, 0.248194664717, -0.419338285923, -0.299710541964, 0.072909116745, -0.0879716128111, 0.0628807023168, 0.101744689047, 0.21888923645, -0.361379235983, 0.245139867067, -0.304993212223, -0.525156974792, 0.000873345532455, 0.480335444212, -0.439277261496, -0.940974533558, -0.2945522964, -0.589957952499, -0.0413432493806, 0.349855303764, 0.0232693050057, -0.584975838661, 0.0588083229959, -0.153540045023, 0.468829095364, 0.615237414837, -0.470928817987, -0.449805498123, 0.374449640512, -0.28412887454, 0.12635101378, -0.116223283112, 0.0188229270279, -0.0936760380864, 0.372985363007, -0.165058523417, 0.104007296264, 0.577144622803, -0.309218585491, -0.0728223621845, -0.0823996290565, -0.514923930168, 0.102164700627, -0.799597203732, 0.0363731980324, 0.550311565399, -0.0905897021294, -0.550099670887, 0.0831759348512, -0.309256374836, -0.323443651199, -0.11093480885, 0.0276359412819, -0.272071838379, -0.100034013391, 0.823453009129, 0.545949459076, -0.0677605867386, -0.160879194736, 0.521573483944, 0.320608735085, -0.112215302885, -0.0835302919149, -0.0640911534429, 0.0208857357502, 0.342691481113, -0.461957722902, 0.43108317256, -0.223896756768, -0.344851404428, -0.593664228916, -0.374976158142, -0.529339671135, -0.340695679188, -0.195017486811, -0.0974618121982, 0.301549136639, -0.333785295486, -0.241046234965, -0.277208000422, -0.297574132681, -0.889755189419, -0.128550425172, -0.180047512054, 0.395120829344, -0.26941677928, 0.0909226536751, 0.201723411679, 0.114779748023, -0.128941982985, -0.336668938398, -0.165393739939, 0.195176973939, -0.115624532104, 0.546708703041, 0.209586113691, -0.0140945501626, -0.00315453624353, -0.205053970218, 0.230554044247, 0.0213234517723, -0.104124419391, -0.301980614662, 0.148923546076, 0.392650425434, -0.251561999321, -0.785662829876, 0.0589303523302, 0.241800203919, 0.0649228617549, 0.128587007523, -0.324615091085, -0.258906751871, 0.473769158125, -0.201317116618, -0.30278018117, 0.459508240223, -0.551006257534, -0.493169516325, -0.566304624081, 0.236484885216, -0.148979380727, -0.389608681202, -0.404490768909, 0.0119635472074, -1.31970310211, -0.556810379028, -0.264738619328, 0.379410684109, -0.0107369152829, -0.356728494167, 0.328995794058, -0.39622810483, 0.538003504276, 0.258066534996, -0.556933164597, 0.346493691206, 1.00652253628, -0.244926318526, -0.139648661017, -0.704957127571, -0.127299875021, 0.951733648777, 0.87039411068, -0.0942110344768, -0.703218221664, 0.391133487225, 0.147731512785, -0.0410939380527, -0.185838386416, -0.0311698466539, 0.242173612118, 0.241510942578, -0.051217790693, -0.319134384394, 0.0297412499785, -0.036984115839, -0.227399155498, -0.169823780656, -0.369953095913, -0.0721863731742, -0.644687533379, -0.16028419137, 0.0497789382935, -0.29839065671, 0.704463064671, -0.227847725153, 0.404548168182, 0.523773372173, -0.294720172882, 0.307150661945, 0.608956992626, 0.521479964256, 0.545867919922, 0.291637718678, -0.0610913224518, -0.49217441678, 0.18885949254, 0.372742950916, 0.180887162685, -0.183727517724, -0.401921272278, -0.194158658385, 0.510446310043, 0.496043473482, -0.456822991371, -0.109792202711, -0.148828849196, 0.471700459719, 0.196458950639, -0.635698974133, 0.11160877347, -0.279748171568, -0.128322467208, 0.136973157525, 0.152544185519, -0.236385956407, -0.503781616688, 0.382612198591, 0.323580741882, 0.142536506057, -0.198670223355, -0.747487068176, 0.261089384556, 0.359869539738, -0.18870177865, -0.249920725822, -0.339892119169, 0.898423433304, -0.259360045195, -0.370633035898, -0.148609548807, -0.52971470356, -0.102137736976, -0.299098163843, -0.411450147629, -0.0975743830204, -0.101265944541, 0.165021672845, -0.182647541165, -0.106156222522, 0.119657792151, -0.208211377263, -0.314904481173, 0.00797762628645, 0.0112475622445, -0.357670992613, 0.404539048672, -0.181079149246, 0.285740882158, 0.202332496643, 0.0608167313039, 0.287276834249, -0.457862734795, 0.309880703688, 0.220650374889, -0.220054715872, 0.138211980462, -0.699477553368, 0.200755447149, 0.589909553528, 0.5602029562, 0.723262965679, -0.55708104372, -0.214356705546, 0.183191120625, -0.233613148332, 0.303535014391, 0.0857863500714, -0.135876402259, 0.0586606673896, 0.379069656134, -0.474004983902, 0.0476555936038, 0.294212818146, 0.119344219565, -0.0216223876923, -0.0461735762656, -0.364545404911, -0.589071631432, -0.912951231003, 0.252641409636, 0.11589153111, -0.159700453281, -0.326051563025, -0.0938815027475, -0.423280090094, -0.341553837061, 0.472398608923, 0.135508596897, 0.170394644141, 0.165682211518, 0.195007026196, -0.367504954338, -0.211165055633, 0.270498007536, -0.345787256956, 0.270992875099, -0.626690804958, 0.0325551852584, -0.188837721944, -0.389090031385, -0.166293501854, 0.0529769212008, -0.118561267853, -0.621582448483, 0.634092748165, 0.0746784806252, 0.143406122923, 0.0433976128697, -0.251817017794, 0.482971280813, 0.589926958084, -0.395296573639, 0.152519837022, -0.0756794661283, -0.442326337099, 0.258955031633, 0.034086432308, 0.148679450154, 0.194083034992, -0.303751885891, 0.347468614578, -0.219399750233, -0.360961973667, -0.101252168417, 0.0643278285861, 0.30460986495, -0.015496795997, -0.317508518696, -0.0715812370181, -0.753070294857, -0.28467848897, -0.0151374032721, 0.0739413350821, 0.173116534948, -0.368360966444, -0.0328403823078, -0.081761687994, 0.339223831892, -0.180474817753, -0.392735540867, 0.313722401857, 0.832724452019, -0.104680746794, -0.367193907499, 0.125167712569, -0.0288922246546, 0.380321443081, 0.355877846479, -0.342632472515, 0.0361616387963, 0.691120743752, -0.131167218089, -0.123408719897, 0.173001527786, 0.362556248903, 0.111326254904, 0.148095831275, 0.276494145393, -0.0422801114619, 0.145789191127, 0.456580221653, 0.687743604183, -0.0635361149907, -0.487026482821, -0.0536798760295, 0.290499329567, -0.257252365351, 0.284238129854, -0.138214468956, 0.271385014057, 0.484819829464, -0.256872206926, -0.447736144066, 0.512245833874, 0.756500661373, 0.411867111921, 0.338565587997, -0.674514770508, -0.305169910192, -0.0578949861228, 0.400421053171, -0.1137194857, 0.189531028271, -0.487585872412, -0.64809679985, -0.0162205863744, 0.134344518185, 0.0696172714233, 0.260535269976, -0.285728007555, -0.0316427461803, -0.670778751373, -0.337795048952, -0.0996651872993, -0.0476932935417, 0.0986134260893, -0.249471321702, -0.00814099423587, 0.0463192090392, 0.421680003405, 0.126967847347, -0.177553460002, 0.385054379702, 0.414351195097, -0.369618564844, -0.0579629801214, 0.021403664723, -0.033271830529, 0.275448292494, 0.374975293875, 0.167479857802, -0.218379050493, 0.279372572899, 0.645205140114, 0.0896425694227, -0.404523164034, -0.352546989918, -0.0341688767076, -0.0414383932948, 0.587391495705, -0.482531934977, 0.0543696880341, -0.331482619047, 0.316195726395, 0.0803859755397, -0.239632353187, -0.167191550136, -0.0157887861133, 0.0669859647751, 0.152309507132, 0.104764387012, 0.100833669305, -0.236839190125, 0.184243693948, 0.12041465193, -0.241180419922, 0.599929630756, -0.111249998212, 0.476123690605, 0.0275491196662, 0.569387614727, -0.434991449118, 0.129588022828, -0.133763894439, -0.3095895648, -0.298461914062, 0.436980426311, -0.113903388381, -0.171226501465, -0.285572528839, -0.150056272745, 0.153229266405, -0.378641486168, 0.668850421906, -0.30952295661, 0.079751804471, 0.434199899435, -0.0930515751243, 0.253315925598, -0.183142960072, -0.0819519758224, -0.293278932571, -0.116167038679, 0.0264510959387, -0.319119006395, -0.307311266661, 0.00481956452131, 0.237418532372, 0.0794789716601, -0.197824135423, -0.357526659966, -0.256382912397, 0.132777735591, 0.318279892206, -0.0531618222594, 0.233718603849, 0.0451786406338, -1.05805635452, -0.313412338495, 0.452723920345, -0.399793833494, -0.148460030556, -0.177501484752, 0.0179707873613, 0.0127085121349, -0.172698259354, -0.425784230232, -0.323390215635, -0.00596148846671, -0.251144081354, 0.675697743893, -0.0884841531515, 0.299249857664, 0.117154799402, -0.354994535446, 0.478085845709, 0.00279094441794, -0.432502776384, -0.0144447721541, 0.160884097219, -0.216258585453, 0.000426440092269, 0.661960244179, -0.242449179292, 0.596243798733, 0.135634630919, 0.0584105923772, -0.528117835522, -0.842608809471, 0.316155642271, 0.2716332376, -0.338095009327, 0.642013311386, -0.433895766735, -0.22602199018, -0.287353634834, -0.0706792771816, -0.186097919941, 0.0319171585143, 0.00208826502785, -0.131310775876, -0.0801445096731, 0.227182686329, 0.0650725141168, 0.310885101557, 0.214585527778, 0.412424534559, 0.226669490337, -0.234037399292, 0.251838862896, 0.369032859802, 0.0231499746442, 0.0203344989568, 0.228693440557, -0.184843987226, -0.223295599222, 0.779653549194, -0.370320439339, 0.25665307045, 0.269827008247, 0.376437872648, -0.404281556606, -0.092129483819, -0.368779033422, 0.483608454466, 0.173235580325, 0.425827860832, 0.232187747955, 0.18199492991, -0.356931865215, -0.0460466444492, 0.208427265286, -0.0528370328248, -0.0866274833679, -0.254204571247, -0.0155757078901, 0.20847697556, -0.44856569171, -0.61966252327, 0.473109811544, -0.196080774069, -0.109102070332, 0.306837946177, -0.484677553177, -0.751867175102, 0.0579593889415, -0.588668525219, 0.748274803162, 0.467081189156, 1.12538015842, -0.390274226665, -0.422624707222, -0.306598633528, 0.709831893444, 0.150826081634, -0.544201254845, 0.639782190323, 0.198120981455, 0.586140692234, 0.0408927612007, -0.264283806086, -0.0716437771916, -0.203621849418, 0.192969188094, 0.111010782421, -0.241321623325, -0.409180402756, -0.383242964745, -0.18724553287, 0.54532033205, 0.31904026866, 0.0199969820678, -0.224509179592, 0.244250103831, -0.0662090182304, -0.165152788162, -0.045413069427, 0.343041598797, -0.163938015699, -0.528438508511, -0.267112582922, 0.309718817472, 0.539167582989, 0.0311706736684, -0.159111067653, -0.742110669613, 0.566340386868, 0.0272132866085, 0.465217143297, -0.291708409786, -0.125376835465, 0.12600903213, 0.0662904083729, 0.233259633183, 0.17283192277, 0.0822020471096, 0.0778817385435, -0.000820057699457, 0.546179831028, -0.208696439862, -0.656307697296, 0.134253963828, -0.209758907557, 0.278120070696, -0.25758394599, 0.0547419860959, -0.372287690639, 0.302204281092, 0.407633006573, 0.205236673355, -0.275920569897, -0.507725775242, 0.124991588295, 0.676386356354, 0.360373944044, -0.529384315014, 0.499229520559, 0.0967811048031, 0.0783446803689, -0.0638069063425, 0.0171148944646, -0.398391872644, 0.532773792744, -0.171478495002, 0.27536419034, 0.188025057316, 0.224178776145, 0.299687176943, 0.425920933485, 0.201258227229, 0.0184365585446, -0.14245171845, -0.114034496248, 0.481977254152, 0.649473726749, 0.313895821571, 0.113943733275, 0.675558567047, -0.00417251931503, -0.461911857128, -0.367138504982, 0.185348391533, 0.522893369198, -0.36487314105, -0.411548554897, 0.21370986104, 0.410751491785, -0.835658490658, -0.666639745235, -0.406063914299, -0.438852399588, -0.281861245632, 0.111891634762, -0.0303278062493, 0.160539522767, -0.234898611903, 0.0998264029622, -0.0865570083261, 0.504289090633, -0.155623584986, -0.631342113018, 0.332029283047, 0.349530756474, -0.618217110634, 0.222039744258, -0.82423555851, -0.828057467937, 0.295978248119, -0.827246844769, 0.0338012427092, -0.240551695228, 0.722697257996, -0.316399693489, 0.613618195057, 0.459160745144, 0.178954139352, -0.214243292809, 0.0607495345175, -0.244216859341, 0.20762732625, 0.241332963109, 0.653276145458, 0.299086630344, 0.106622710824, -0.204596489668, -0.062441252172, 0.15427698195, 0.133646249771, 0.148002400994, -0.373497158289, -0.290177345276, 0.126465275884, -0.412142604589, -0.47807008028, 0.628624022007, 0.0914865732193, 0.0323539301753, -0.173378914595, -0.469193726778, -0.828047573566, -0.361548185349, -0.168318152428, -0.323063850403, -0.124064549804, 0.0582916922867, 0.438749343157, 0.68442940712, -0.300360709429, 0.294909328222, 0.128743678331, 0.348173558712, -0.228440105915, -0.0270835030824, -0.433656573296, 0.369668841362, -0.308600515127, 0.365284055471, -0.673052966595, -0.189704492688, -0.0530697032809, -0.0168927740306, -0.274845063686, -0.199872329831, 0.182394474745, -0.819949150085, 0.226003870368, -0.227072358131, -0.361823171377, -0.160746246576, -0.223619773984, -0.670349121094, -0.156297490001, -0.573216974735, 0.248604759574, -0.542316496372, 0.368392884731, 0.142644420266, 0.147660300136, -0.0473959781229, -0.142904952168, -0.497010856867, 0.157088115811, 0.330286502838, 0.183208569884, -0.0248327944428, -0.1519010216, 0.088474586606, 0.205937057734, -0.136989936233, 0.27269411087, 0.508547544479, 0.135484799743, -0.293869286776, -0.041024826467, -0.00393711589277, -0.220801159739, 0.309434950352, -0.520574510098, -0.299849301577, -0.56118786335, -0.691200375557, -0.23916245997, -0.105003371835, 0.238450333476, -0.321029871702, -0.172825857997, -0.973243355751, 0.0147989718243, 0.673430562019, 0.561264395714, -0.66096585989, -0.0373016521335, 0.240049481392, 0.147330850363, 0.255987077951, 0.290556788445, -0.663549125195, 0.262611269951, -0.233406290412, 0.2513474226, 0.253098130226, 0.136670470238, -0.144896656275, 0.364002764225, 0.975185513496, -0.0405109785497, -0.448907256126, 0.460708618164, 0.272006630898, 0.277762204409, 0.0219166465104, -0.448736727238, -0.0534540005028, -0.784090459347, 0.470430612564, -0.0264166090637, -0.904743492603, -0.350075691938, 0.41157707572, 0.150189146399, -0.117259636521, 0.17916546762, -0.695175230503, 0.379684358835, -0.0881545096636, 0.373303085566, 0.0430721864104, 0.0381501764059, 0.477151572704, 0.00988717470318, -0.0436263419688, -0.422077775002, -0.0866391733289, -0.603933274746, 0.148453131318, 0.221070036292, -0.108046241105, -0.303051143885, 0.441788882017, -0.00440698349848, 0.365677624941, 0.293478250504, -0.270659476519, -0.319292187691}
+};
+std::vector< std::vector<float> > U_o = {
+{-0.230232179165, 0.0205486081541, -0.401903271675, 0.109920702875, -0.691217064857, 0.0470181033015, 0.133465856314, -0.173473194242, -0.0576262511313, 0.183640822768, 0.405053257942, 0.24936440587, -0.594521760941, -0.718641936779, -0.226488083601, -0.299191623926, -0.39187514782, 0.345947921276, -0.218880310655, 0.285045921803, -0.436422526836, -0.0438544116914, -0.0655047297478, -0.130614161491, 0.301144987345, 0.572843015194, 0.159976229072, -0.451845765114, -0.365841031075, -0.810667037964, -0.140651345253, 0.00223093712702, -0.0169935990125, 0.716104567051, 0.487425595522, -0.140877991915, 0.646854639053, -0.339587509632, 0.174669191241, -0.483658313751, 0.2889328897, -0.132459461689, 0.220022514462, 0.555271863937, -0.141440182924, 0.0555354841053, -0.179256826639, 0.289142340422, -0.434483736753, -0.10009033978, -0.495820224285, 0.512155056, -0.356727391481, -0.211471363902, 0.421799659729, 0.234898149967, -0.706707239151, -0.448694169521, 0.659977853298, -0.0469280704856, 0.0946899652481, -0.586544394493, 0.139435768127, 0.38280826807, 0.0721923559904, 0.186812922359, -0.185512423515, 0.72023499012, -0.485877066851, -0.0822856649756, 0.212789356709, 0.218568772078, -0.266736507416, -0.607727885246, 0.000461816671304, -0.0694766789675, 0.0195119883865, -0.264453053474, -0.298973470926, -0.93114644289, -0.130476161838, 0.056644808501, -0.169811531901, -0.146569475532, 0.423955261707, -0.457307547331, 0.142826199532, 0.321131616831, -0.180789917707, -0.0959079638124, 0.22587376833, -0.0663760453463, -0.164583325386, -0.140932574868, -0.199941024184, -0.181526333094, -0.0203421991318, 0.195820048451, -0.157560244203, 0.446546345949, 0.0922764390707, 0.28920724988, -0.0416567362845, 0.288586825132, 0.784974098206, -0.604826271534, 0.00307735870592, -0.00698337610811, -0.303714960814, 0.0230147317052, -0.486879169941, -0.162032619119, 0.368861049414, -0.259443819523, 0.407054841518, 0.148995965719, -0.416047304869, 0.259648382664, 0.28815099597, 0.762510418892, -0.902653813362, 0.440671890974, 0.629763305187, 0.335902243853, 0.835532844067, 0.0369899496436, 0.17930842936, 0.0786409452558, 0.106598071754, 1.23513257504, 0.616916179657, 0.29204210639, -0.0242422111332, -0.153707757592, 0.127918958664, -0.119772262871, 0.544666051865, 0.324779838324, 0.239006921649, -0.0924495905638, -0.607812285423, 0.769949734211, 0.601655304432, -0.0130537552759, -0.583716511726, -0.686632812023, 0.483102440834, -0.0151197090745, 0.709682703018, 0.758449852467, -0.140919730067, 0.293156296015, -0.537789642811, 0.63957130909, -0.200582161546, 0.133771359921, 0.4561406672, -0.0509933903813, 0.199158996344, -0.603383898735, 0.13383051753, -0.958775579929, 0.19699332118, -0.0974606499076, 0.317563951015, 0.680873453617, 0.212780579925, -0.0407421551645, -0.425788760185, 0.396682858467, 0.385065585375, 0.283968031406, -0.964271366596, 0.218730434775, 0.482916682959, 0.332988202572, -0.223672240973, -0.032158844173, -0.260298401117, -1.68755960464, 0.441846311092, 0.0651218891144, -0.226209715009, -0.291343122721, 0.260928273201, 0.0178945269436, 0.823356032372, 0.00574315525591, 0.00569802382961, 0.332162946463, -0.484819918871, 0.268787294626, 0.21328894794, 0.053284380585, -1.03840112686, 0.158958375454, 0.624661266804, 0.436532795429, 0.038346990943, -0.197258070111, 0.108011722565, 0.334964424372, 0.452759444714, -0.0072251195088, 0.14552937448, 1.26336550713, 0.434452921152, -0.25795352459, -0.0116482712328, 0.085065074265, -0.738467156887, 0.120552845299, 0.115385010839, -0.215058803558, -0.497080981731, -0.138640195131, -0.924027681351, -0.048629976809, 0.269809126854, 0.211867496371, 0.230882197618, 0.0495937131345, -0.358516395092, -0.371314316988, -0.188390687108, -0.115469604731, 0.486744374037, -0.210745409131, 0.0571626313031, -0.405347138643, -0.0724542587996, -0.125774487853, 0.214891970158, -0.182782143354, -0.497832238674, -0.15929967165, -0.0369414910674, -0.108604483306, 0.111244261265, -0.489217430353, -0.528547942638, 0.186059564352, -0.00850142166018, 0.0436340086162, 0.335779368877, 0.145701736212, -0.685704827309, 0.378927499056, 0.814988493919, 0.0138626499102, -0.371524035931, 0.274067401886, 0.679609954357, 0.483737826347, 0.131814718246, 0.351422905922, 0.537636816502, 0.174367129803, 0.0121098076925, 0.246450707316, 0.17421156168, 0.284833520651, 0.142588481307, 0.407764256001, 0.307735145092, 0.337948232889, 0.520263254642, 0.479906588793, 0.417292118073, 0.317961573601, 0.369405925274, 0.383081316948, -0.141126438975, 0.269783645868, 0.324794441462, 0.174264892936, -0.150778159499, 0.660377502441, -0.393305420876, 0.484085559845, -0.937421798706, -0.156022235751, -0.31157925725, 0.00634649256244, 0.114026561379, -0.0178149919957, 0.306646317244, -0.52957957983, 1.02926778793, -0.50990998745, 0.968846261501, 0.0998627468944, -0.322920173407, 0.189471021295, -0.117565631866, 0.455198228359, -0.0231772474945, 0.0894395112991, -0.105348929763, 0.61057806015, -0.236860319972, -1.05028641224, 0.395838499069, -0.662855446339, -0.0940272808075, -1.11867642403, 0.390377759933, -0.0936558395624, -0.556634426117, -0.666942358017, -0.466880619526, 0.3956579566, 0.499339640141, 0.280326336622, 0.386395812035, 2.36877240241e-05, -0.497093200684, 0.075920149684, 0.120664410293, 0.102253749967, -0.0604435428977, -0.29989272356, 0.192053884268, 0.0875131040812, 0.184831604362, 0.837235987186, -0.137301489711, 0.510612308979, -0.190522491932, -0.0028200247325, 0.0143071198836, -0.0160747561604, -0.274926126003, 0.236399829388, -0.602124333382, 0.336574524641, -0.302964925766, 0.800043821335, -0.154450595379, 0.317300826311, 0.488817811012, 0.108004659414, 0.375342041254, -0.172089263797, 0.386206269264, -0.219486773014, 0.310803622007, 0.12701100111, -0.0349657200277, 0.621266067028, 0.596777200699, -0.078187443316, -0.355752080679, -0.0467068068683, 0.0521030016243, 0.114057518542, -0.0252706017345, 0.150653958321, -0.0470893755555, 0.0066180867143, -0.0278404839337, 0.237126126885, 0.810919880867, -0.0553170219064, -0.146957442164, 0.489920407534, -0.301968842745, 0.598005890846, 0.113271549344, 0.436753541231, -0.714772462845, 0.345493525267, -0.0359437949955, 0.174665823579, 0.378943115473, 0.021457279101, -0.426008880138, 0.0124987950549, -0.526060521603, 0.200031235814, -1.24587607384, -0.597509741783, 0.714687824249, 0.871401548386, -0.231417968869, 0.0194307602942, 0.33912011981, -0.290394037962, 0.136520594358, 0.362242549658, -0.245033979416, 1.09198760986, 0.323082149029, 0.196922659874, -0.474113762379, -0.154468342662, -0.441907376051, -0.59861344099, 0.0561420880258, 0.380189597607, 0.377747148275, -0.255721896887, 0.180880144238, 0.362433642149, 0.721752762794, 0.838150918484, 0.182588756084, 0.0935470685363, 0.305843383074, -0.324756771326, -0.310819387436, -1.16464281082, -0.290249615908, -0.50999546051, 0.283817321062, -0.434470951557, -0.126545652747, -0.295378774405, 0.565843939781, -0.807359278202, -1.00800383091, 0.197524830699, -0.442735254765, -0.749307096004, -0.769624471664, 0.456985831261, -1.21000897884, 0.0859573632479, -0.0873735621572, -0.112147927284, -0.301290512085, -0.171894684434, 0.281311690807, -1.19254004955, 0.0904107093811, -0.290790945292, -0.145752802491, -0.19665145874, -0.284464478493, 0.771230578423, 0.373421400785, 0.182227656245, 0.180574789643, -0.132402479649, 0.246747225523, 0.0957783013582, 0.390556126833, -0.0179129913449, -0.0157641544938, -0.394919961691, 0.132176667452, 0.350833982229, 0.152620702982, 0.100988700986, 0.0881125405431, -0.156632050872, 0.400717109442, 0.521214842796, 0.470715820789, -0.220873698592, -0.294844329357, -0.205065280199, -0.196818709373, 0.795477926731, -0.384722739458, -1.06169712543, -0.167890742421, 0.49599763751, -0.0101557439193, -0.558584392071, -0.702517211437, -0.179195135832, -0.0889069065452, 1.24287176132, 0.386941492558, -0.377963095903, -0.740742206573, 0.127455756068, -0.230323076248, -0.163900479674, -0.395755589008, 0.43536195159, 0.738106548786, -0.799264013767, -0.20603042841, 0.710767865181, 0.523515284061, -0.907929956913, 0.0788556262851, 0.111601270735, -0.0745002478361, 0.346990317106, 0.32616391778, -0.23171889782, -0.473845392466, 0.664731264114, 0.0312324296683, -0.293554157019, -0.0703734606504, -0.184768974781, -0.0756682902575, 0.371084541082, 0.0839636474848, 0.0920626595616, -0.516082465649, -0.563040733337, 0.143838390708, 0.293257385492, 0.390136986971, -0.088386811316, 0.0194950494915, -0.309329986572, 0.0987806171179, 0.255824953318, -0.00011260494648, -0.172500222921, 0.112487077713, -0.129405960441, 0.31120377779, 0.322653442621, 0.397172451019, -0.347260117531, -0.409591943026, -0.172383740544, -0.172353431582, -0.104770876467, -0.502486109734, -0.258416891098, -0.239610090852, 0.695222616196, -0.130661696196, 0.806845724583, -0.206436693668, -0.744152188301, 0.464515298605, 0.818568170071, -0.438517779112, -0.752291917801, -0.104600019753, 0.0779007822275, -0.0495877750218, 0.341687679291, 0.0815256163478, -0.0824779421091, 0.3301846385, -0.296553939581, -0.439097911119, 0.135820060968, 0.797622740269, -0.219913393259, -0.220496103168, -0.109801113605, 0.599159359932, 0.102990858257, 0.608977615833, 0.0504399836063, -0.493425905704, 0.0287420041859, -0.0884666144848, 0.184674471617, -0.197273418307, 0.159897565842, -0.428218573332, -0.0391625054181, 0.115218162537, -0.210350200534, -0.492515921593, -0.2302210778, 0.249826833606, 0.0847941711545, -0.356171488762, -0.401976972818, -0.254000753164, 0.217901006341, -0.231287255883, 0.341797351837, -0.154923707247, 0.372051864862, -0.210665076971, 0.115214712918, -0.1160717085, -0.352732926607, 0.238445043564, 0.320423394442, 0.0782100856304, -0.153607711196, -0.021156443283, -0.86895608902, -0.0472883842885, 0.405406236649, -0.43440887332, -0.541340589523, -0.205899387598, 0.571096301079, 0.527008593082, 0.229026332498, -0.0568662919104, -0.521931231022, -0.756743788719, -0.060257948935, 0.454650580883, 0.392847001553, 0.514076173306, -0.195735380054, -0.149251639843, 0.0240337196738, -0.32173460722, -0.00516948476434, 0.149106442928, 0.327200859785, -0.116163298488, 0.0189112257212, -0.192452773452, 0.389188468456, -0.52135938406, -0.404690414667, 0.603824496269, 0.219251781702, 0.243884593248, -0.28382691741, -0.290861874819, -0.477886646986, 0.0771175771952, -0.310908734798, -0.325133025646, 0.464059889317, 0.328977435827, 0.129672557116, -0.553050339222, -0.507432222366, 0.115787640214, 0.405775815248, -0.0289540793747, -0.214790910482, 0.00926195364445, -0.401621758938, 0.0141945555806, 0.695082128048, -0.306836396456, -0.280148535967, 0.191328331828, -0.0740591660142, 0.237794160843, 0.521210551262, 0.507326960564, -0.326399624348, -0.282526016235, -0.453115671873, -0.323135435581, -0.684462785721, -0.659346997738, -0.275617092848, 0.151040226221, 0.233532860875, -0.0550286248326, 0.241681188345, -0.240301191807, -0.274218916893, -0.350826144218, -0.604172348976, 0.211623609066, -0.360034793615, -0.295849204063, -0.0619548112154, -0.042968723923, -0.0480784736574, 0.299180865288, -0.148937031627, -0.901235699654, -0.101308859885, -0.394019365311, -0.209546416998, 0.495187491179, -0.454454779625, -0.249904125929, -0.13583612442, -0.555691540241, 0.370698779821, -0.5843744874, 0.251811653376, 0.0172328464687, 0.457344740629, -0.283511608839, -0.0333109535277, 1.56934905052, -0.00757428444922, -0.0676159262657, 0.0920815989375, -0.482059985399, 0.372551232576, -0.43467310071, 0.742170333862, -0.0253379456699, 1.08074021339, -0.698093533516, -0.619763255119, 0.420928686857, -0.159713298082, 0.677888333797, 0.957087218761, -0.198817029595, -0.216832309961, 0.721533238888, 0.408280283213, -0.17367580533, 0.637643396854, 0.246913656592, 0.927109181881, -0.452713102102, 1.52946841717, 0.0325671993196, -0.219425037503, 1.21932530403, 0.0897421017289, -0.216693207622, 0.166701167822, -0.117182388902, -0.317105710506, -0.33998015523, 0.203328862786, -0.15231962502, -0.0514479577541, 0.385309845209, 0.737704575062, 0.353906542063, -0.291469246149, -0.510806381702, -0.0358709841967, 0.379145056009, 0.120908439159, -0.053712785244, -0.215732142329, -0.32273119688, -0.391137242317, -0.096145927906, -0.0706685110927, 0.439254701138, 0.144891545177, -1.15791594982, -0.476177155972, -0.386640161276, -0.0696934089065, 0.00766070419922, 0.0915936976671, -0.194772407413, 0.0288319084793, 0.807722628117, -0.312456071377, 0.024683419615, -0.134210512042, -0.658850967884, -0.444511920214, 1.16682302952, 0.412955373526, -0.86106300354, -0.843534350395, -0.0756809189916, 0.249314174056, -0.768764734268, -0.18570099771, -0.64992171526, -1.06809818745, -0.241304412484, -1.09838747978, -0.408396184444, -0.448166787624, 0.306912541389, -0.644407331944, 0.264287084341, 0.352415382862, -0.32395285368, -0.11422291398, 0.169177040458, -0.171893030405, -0.256403177977, -0.151077628136, 0.51526093483, -0.780808508396, -0.556867837906, 0.540703356266, -0.94756937027, 0.518447637558, -0.379737138748, 0.325614154339, -0.589700877666, -0.121931925416, -0.376903027296, 0.674576640129, 1.0511289835, -0.200893417001, 0.866664469242, -0.31772223115, 0.389687240124, 0.329201370478, 0.14609798789, -0.0465747490525, -0.130747646093, 1.11257493496, -0.171550154686, -1.10172557831, 0.783048510551, 0.654971837997, 0.430614262819, 0.0185522232205, 0.533750295639, 0.957171976566, -0.411320418119, 0.0443305559456, 0.879363298416, 0.0413797199726, 0.147208213806, -0.547723352909, -0.380599319935, 0.0328748337924, 0.162938177586, 0.309380620718, -0.144598171115, -0.365523636341, 0.12974075973, 0.051276486367, 0.318579405546, 0.301273941994, -0.0293224602938, -0.539349555969, -0.0128162642941, -0.396678030491, -0.154957085848, 0.207027703524, -0.271222531796, -0.338817328215, -0.353360652924, 0.0440826974809, 0.557435572147, -0.364424318075, -0.408349364996, -0.255339831114, 0.610197246075, 0.3794259727, -0.00990682188421, -0.0818183645606, 0.212990865111, 0.405529648066, -0.532237768173, 0.572834432125, 0.183916419744, -0.864466190338, -0.089433401823, 0.35382142663, 1.23338615894, -0.138091757894, -0.385909616947, 0.218497782946, 0.603306472301, 0.558208048344, -0.237068578601, 0.642093062401, 0.765398025513, 0.238009110093, 0.127825573087, -0.487903863192, 0.0951134040952, -0.262359976768, -0.606513559818, 0.107818476856, 0.220998287201, 0.115167543292, -0.346906840801, -0.308289676905, -0.196890816092, 1.19753932953, -0.632979691029, 0.350641727448, 0.118041612208, 0.115423731506, -0.767681896687, -1.1139986515, -0.168996348977, 0.203874543309, -0.0669453367591, -0.168869003654, -0.289925992489, -0.0373433195055, -0.785607755184, -0.120450884104, -0.991163671017, -1.64675581455, -0.0369265303016, -0.15207824111, -0.585144102573, -0.652086734772, -0.156884670258, -1.01922428608, -0.558863282204, -0.0643538758159, 0.0412841588259, 0.339178562164, 0.303961127996, -0.376203209162, 0.448310941458, -0.186875894666, 0.181232169271, -0.402108848095, 0.672553420067, -0.200394809246, 0.117735303938, 0.0478924252093, -1.02634394169, -0.210108563304, -0.256260812283, 0.0259675178677, -0.355624139309, 0.039109505713, -0.247299253941, -1.17434430122, 0.278623580933, -1.04232430458, 0.0840487703681, 0.240674853325, -0.695171833038, -0.736045897007, -1.0585244894, -0.282407969236, -0.0547968931496, -0.0205924920738, -0.454886257648, -0.783143937588, 0.0938883125782, 0.0397834517062, -0.166149884462, -0.0343275517225, 0.0802892446518, -0.211236014962, 0.200299948454, 0.164131179452, -1.26071095467, 0.380297392607, -0.10085054487, -0.0734983608127, 0.428060978651, 0.474595427513, 0.70434564352, 0.207352936268, -0.149105146527, -0.274008482695, -0.405232489109, -0.099853143096, 0.637039959431, -0.853150248528, 0.143071815372, 0.891714990139, 0.745768725872, 0.383694887161, -0.741264939308, 0.061810079962, 0.451800614595, -0.658157885075, -0.523318350315, -0.301910966635, 0.422813117504, -0.749253630638, -0.375865846872}
+};
+std::vector< std::vector<float> > b_i = {
+{-0.0646453052759, -0.833576738834, -0.0266484506428, 0.191880121827, 0.17440316081, 0.258504778147, 0.304691702127, -0.135863363743, 0.421445906162, -0.235500946641, -0.360742539167, -0.154070556164, -0.435084909201, -0.0338653549552, -0.240953952074, 0.0292105004191, -0.378072559834, -0.146301329136, -0.268983244896, -0.142332673073, 0.277948409319, -0.192356497049, -0.415451616049, -0.453225821257, -0.351215183735, -0.100444279611, -0.369603067636, 0.113715209067, 0.300046652555, -0.408928126097, 0.599760174751}
+};
+std::vector< std::vector<float> > b_f = {
+{0.700906693935, 0.563576161861, 1.37404668331, 0.664588212967, 0.655001282692, 1.50752604008, 0.901046395302, 0.908896207809, 0.531699061394, -0.175342053175, 0.622060060501, 0.766630768776, 0.517459094524, 0.655570209026, 0.967856347561, 0.531598210335, 0.995056331158, 0.863079667091, 0.805407762527, 1.07394945621, 1.33595705032, 0.889388442039, 0.709481477737, 0.916946470737, 0.724635124207, 0.30798125267, 0.742200016975, 0.249754697084, 0.470162123442, 0.843433737755, 0.515080571175}
+};
+std::vector< std::vector<float> > b_c = {
+{0.111251153052, -0.318186342716, -0.178570881486, -0.00616357009858, 0.0236715469509, -0.0270600337535, -0.447214305401, 0.350495159626, -0.597518026829, -0.242235258222, -0.0601634345949, 0.00118016276974, 0.448368012905, -0.0442144423723, -0.406291365623, 0.229264989495, -0.302716135979, 0.554887712002, 0.35920727253, 0.304902613163, -0.0852948501706, -0.0758324190974, -0.235283449292, -0.319113492966, -0.173131242394, -0.12002632767, -0.052010346204, -0.0521944798529, -0.00280740624294, 0.175736755133, -0.111001990736}
+};
+std::vector< std::vector<float> > b_o = {
+{-0.0560160428286, 0.0134335802868, -0.212648883462, 0.247949317098, 0.192606985569, -0.326273173094, -0.41331744194, -0.304275363684, 0.279461324215, -0.530653178692, -0.315937668085, -0.477790594101, -0.590563416481, 0.314803212881, -0.128587067127, -0.0400473587215, -0.361676484346, -0.267935723066, -0.186445727944, -0.36678853631, -0.281785964966, -0.122770354152, -0.110378704965, -0.724780499935, -0.183713123202, 0.106771074235, -0.233618751168, 0.157023027539, 0.219069510698, -0.396034121513, 0.614786446095}
+};
diff --git a/coeff.bin b/coeff.bin
new file mode 100644
index 0000000..eb40001
--- /dev/null
+++ b/coeff.bin
diff --git a/config.cpp b/config.cpp
new file mode 100644
index 0000000..ac0f1d3
--- /dev/null
+++ b/config.cpp
@@ -0,0 +1,19 @@
+#include <vector>
+#include <string>
+
+#include "config.h"
+
+namespace config{
+  std::vector<std::string> LABELS_NAMES = {
+     "Pushup",
+     "Pushup_Incorrect",
+     "Squat",
+     "Situp",
+     "Situp_Incorrect",
+     "Jumping",
+     "Lunge"
+   };
+
+   int N_FEATURES = 2;
+   int SEGMENT_TIME_SIZE = 1;
+}
diff --git a/config.h b/config.h
new file mode 100644
index 0000000..2ab3ac1
--- /dev/null
+++ b/config.h
@@ -0,0 +1,26 @@
+/*
+Config file with global constants.
+*/
+
+#ifndef CONFIG_H_
+#define CONFIG_H_
+
+#include <vector>
+#include <string>
+
+namespace config{
+  // Labels (activities) that are considered by the model
+  extern std::vector<std::string> LABELS_NAMES;
+
+  // Number of features (3 if only acceleration is used)
+  // 9 if acceleration + gyro + magnetometer readings are used
+  extern int N_FEATURES;
+
+  // Size of a sliding window (how many samples does it contain)
+  extern int SEGMENT_TIME_SIZE;
+
+  // Sliding window shift (number of samples)
+  extern int TIME_STEP;
+}
+
+#endif
diff --git a/logs/normal45-270-v100-with-friction2-iter10-15.log b/logs/normal45-270-v100-with-friction2-iter10-15.log
new file mode 100644
index 0000000..13a886e
--- /dev/null
+++ b/logs/normal45-270-v100-with-friction2-iter10-15.log
diff --git a/lstm_infer.cpp b/lstm_infer.cpp
new file mode 100644
index 0000000..2ad5d60
--- /dev/null
+++ b/lstm_infer.cpp
@@ -0,0 +1,32 @@
+/*
+Compile with:
+g++ --std=c++11 -Wall -O3 test_cpp_model.cpp onnx_model.cpp utils.cpp config.cpp -o output
+*/
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <malloc.h>
+
+#include <iostream>
+#include "onnx_model.h"
+#include "config.h"
+
+OnnxModel *model;
+void lstmSetup(void)
+{
+  model = new OnnxModel;
+  // Initialize model.
+  model->LoadModel();
+}
+
+void runLstm(double in1, double in2, double *out1, double *out2)
+{
+  Tensor in(config::SEGMENT_TIME_SIZE, config::N_FEATURES);
+  Tensor out;
+    
+  in.data_.at(0) = in1;
+  in.data_.at(1) = in2;
+  model->Apply(&in, &out);
+  *out1 = out.data_.at(0);
+  *out2 = out.data_.at(1);
+}
diff --git a/mk.sh b/mk.sh
new file mode 100755
index 0000000..79d6365
--- /dev/null
+++ b/mk.sh
@@ -0,0 +1 @@
+g++ --std=c++11 -Wall -O3 motor-pdm.cpp psensors.cpp lstm_infer.cpp onnx_model.cpp config.cpp -lpthread -o motpdm
diff --git a/models/configheader.py b/models/configheader.py
new file mode 100644
index 0000000..21876f4
--- /dev/null
+++ b/models/configheader.py
@@ -0,0 +1,66 @@
+"""
+Config file.
+All other scripts import variables from this config
+to their global namespace.
+"""
+##################################################
+### GLOBAL VARIABLES
+##################################################
+COLUMN_NAMES = [
+    'activity',
+    'acc-x-axis',
+    'acc-y-axis',
+    'acc-z-axis',
+    'gyro-x-axis',
+    'gyro-y-axis',
+    'gyro-z-axis',
+    'mag-x-axis',
+    'mag-y-axis',
+    'mag-z-axis'
+]
+
+LABELS_NAMES = [
+    'Pushup',
+    'Pushup_Incorrect',
+    'Squat',
+    'Situp',
+    'Situp_Incorrect',
+    'Jumping',
+    'Lunge'
+]
+
+# Data directories
+DATA_PATH = 'data/data.pckl'
+
+# Model directories
+MODEL_PATH = 'models/model.h5'
+MODEL_PATH_DIR = 'models/'
+HEADERFILE_NAME = 'LSTM_model.h'
+
+##################################################
+### MODEL
+##################################################
+# Used for shuffling data
+RANDOM_SEED = 13
+
+# Model
+N_CLASSES = len(LABELS_NAMES)
+N_FEATURES = 3
+
+# Hyperparameters
+N_EPOCHS = 30
+LEARNING_RATE = 0.0005
+N_HIDDEN_NEURONS = 30
+BATCH_SIZE = 30
+DROPOUT_RATE = 0.2
+
+##################################################
+### DATA COLLECTION/PREPROCESSING
+##################################################
+
+# Data preprocessing
+TIME_STEP = 5
+SEGMENT_TIME_SIZE = 40
+
+# Train/test proportion
+TEST_SIZE = 0.2
diff --git a/models/make_header.py b/models/make_header.py
new file mode 100644
index 0000000..36cb0ee
--- /dev/null
+++ b/models/make_header.py
@@ -0,0 +1,305 @@
+"""
+Script for making a C++ headerfile with LSTM network
+based on an ONNX model placed in MODEL_PATH
+(see config file).
+"""
+
+import sys
+_b=sys.version_info[0]<3 and (lambda x:x) or (lambda x:x.encode('latin1'))
+import os
+import numpy as np
+import pandas as pd
+from configheader import *
+from preprocessingheader import get_convoluted_data
+import onnx
+from onnx import numpy_helper
+import argparse
+
+LAYER_DENSE = 1
+LAYER_ACTIVATION = 2
+LAYER_LSTM = 3
+
+ACTIVATION_LINEAR = 1
+ACTIVATION_RELU = 2
+ACTIVATION_SOFTPLUS = 3
+ACTIVATION_SIGMOID = 4
+ACTIVATION_TANH = 5
+ACTIVATION_HARD_SIGMOID = 6
+
+"""
+Given activation return a number assigned to that activation.
+"""
+def activation_type(activation):
+    if activation == 'linear':
+        return ACTIVATION_LINEAR
+    elif activation == 'relu':
+        return ACTIVATION_RELU
+    elif activation == 'softplus':
+        return ACTIVATION_SOFTPLUS
+    elif activation == 'tanh':
+        return ACTIVATION_TANH
+    elif activation == 'sigmoid':
+        return ACTIVATION_SIGMOID
+    elif activation == 'hard_sigmoid':
+        return ACTIVATION_HARD_SIGMOID
+
+"""
+Append functions are used to append strings to the "file",
+so that it has a C/C++ headerfile style.
+"""
+def append_vector(file, type, name, vector_object):
+    file.write("std::vector<{}> {} = {{\n".format(type, name))
+    file.write(', '.join(map(str, vector_object.tolist())))
+    file.write("};\n\n")
+
+def append_list(file, type, name, list):
+    file.write("std::vector<{}> {} = {{".format(type, name))
+    file.write(', '.join(map(str, list)))
+    file.write("};\n")
+
+def append_vector_of_vectors(file, n, type, name, array):
+    file.write("std::vector< std::vector<{}> > {} = {{\n".format(type, name))
+    for i in range(n):
+        append_numbers(file, array[i].flatten())
+        if(i != n - 1):
+            file.write(",\n")
+    file.write("\n};\n")
+
+def append_numbers(file, array):
+    file.write("{")
+    file.write(', '.join(map(str, array.tolist())))
+    file.write("}")
+
+def append_includes(file):
+    file.write("/*This is the header file generated from the ONNX model*/\n\n")
+    file.write("#include <string>\n")
+    file.write("#include <vector>\n")
+    file.write("\n")
+
+"""
+Function used for creating and saving to disk a headerfile.
+Input: ONNX model
+Output: None (saves file to disk)
+"""
+def make_header(model, lstm_layers_count):
+      # Count number of each layer type
+      dense_layers_count = 1
+      num_layers = dense_layers_count + lstm_layers_count
+      INIT_Weights = model.graph.initializer
+
+      if os.path.exists(HEADERFILE_NAME):
+        os.remove(HEADERFILE_NAME)
+
+      with open(HEADERFILE_NAME, "a") as file:
+        append_includes(file)
+        file.write("unsigned int NUM_LAYERS = " + str(num_layers) + ";\n")
+        # LSTM layer
+        recurrent_activation_list = []
+        activation_list = []
+        return_sequences_list = []
+
+        num_units_list = []
+
+        W_i_list = []
+        W_f_list = []
+        W_c_list = []
+        W_o_list = []
+
+        U_i_list = []
+        U_f_list = []
+        U_c_list = []
+        U_o_list = []
+
+        b_i_list = []
+        b_f_list = []
+        b_c_list = []
+        b_o_list = []
+
+        W_i_ROWS_list = []
+        W_i_COLS_list = []
+        W_f_ROWS_list = []
+        W_f_COLS_list = []
+        W_c_ROWS_list = []
+        W_c_COLS_list = []
+        W_o_ROWS_list = []
+        W_o_COLS_list = []
+        U_i_ROWS_list = []
+        U_i_COLS_list = []
+        U_f_ROWS_list = []
+        U_f_COLS_list = []
+        U_c_ROWS_list = []
+        U_c_COLS_list = []
+        U_o_ROWS_list = []
+        U_o_COLS_list = []
+        b_i_SHAPE_list = []
+        b_f_SHAPE_list = []
+        b_c_SHAPE_list = []
+        b_o_SHAPE_list = []
+
+        # Fully connected
+        weights = np.transpose(numpy_helper.to_array(INIT_Weights[5]))
+        biases  = np.transpose(numpy_helper.to_array(INIT_Weights[6]))
+        activation = 'linear'
+        print "FC weights:" + str(weights.shape)
+        print "FC biases :" + str(biases.shape)
+
+        file.write("unsigned int DENSE_WEIGHTS_ROWS = " + str(weights.shape[0]) + ";\n")
+        file.write("unsigned int DENSE_WEIGHTS_COLS = " + str(weights.shape[1]) + ";\n")
+        file.write("unsigned int DENSE_BIASES_SHAPE = " + str(biases.shape[0]) + ";\n")
+        file.write("unsigned int DENSE_ACTIVATION = "   + str(activation_type(activation)) + ";\n\n")
+
+        append_vector(file, "float", "DENSE_WEIGHTS", weights.flatten())
+        append_vector(file, "float", "DENSE_BIASES", biases)
+
+        # LSTM layer attriutes
+        recurrent_activation = 'sigmoid'
+        activation = 'sigmoid'
+        return_sequences = 0
+
+        recurrent_activation_list += str(activation_type(recurrent_activation))
+        activation_list += str(activation_type(activation))
+        return_sequences_list += str(return_sequences)
+
+        Wonnx = numpy_helper.to_array(INIT_Weights[0])
+        Uonnx = numpy_helper.to_array(INIT_Weights[1])
+        bonnx = numpy_helper.to_array(INIT_Weights[2])
+        W   = np.transpose(Wonnx[0,:])
+        U   = np.transpose(Uonnx[0,:])
+        bIN   = np.transpose(bonnx[0,:])
+
+        b = bIN[:len(bIN)/2]
+        print "bIN shape " + str(bIN.shape)
+        
+        num_units = int(int(W.shape[1])/4)
+        num_units_list.append(num_units)
+        # file.write("unsigned int N_HIDDEN_NEURONS = " + str(num_units) + ";\n\n")
+        print "W  shape:" + str(W.shape)
+        print "U  shape:" + str(U.shape)
+        print "b  shape:" + str(b.shape)
+        W_i = W[:, :num_units]
+        W_o = W[:, num_units: num_units * 2]
+        W_f = W[:, num_units * 2: num_units * 3]
+        W_c = W[:, num_units * 3:]
+        print "Wi shape:" + str(W_i.shape)
+        print "Wf shape:" + str(W_f.shape)
+        print "Wc shape:" + str(W_c.shape)
+        print "Wo shape:" + str(W_o.shape)
+
+        U_i = U[:, :num_units]
+        U_o = U[:, num_units: num_units * 2]
+        U_f = U[:, num_units * 2: num_units * 3]
+        U_c = U[:, num_units * 3:]
+        print "Ui shape:" + str(U_i.shape)
+        print "Uf shape:" + str(U_f.shape)
+        print "Uc shape:" + str(U_c.shape)
+        print "Uo shape:" + str(U_o.shape)
+
+        b_i = b[:num_units]
+        b_o = b[num_units: num_units * 2]
+        b_f = b[num_units * 2: num_units * 3]
+        b_c = b[num_units * 3:]
+        print "Bi shape:" + str(b_i.shape)
+        print "Bf shape:" + str(b_f.shape)
+        print "Bc shape:" + str(b_c.shape)
+        print "Bo shape:" + str(b_o.shape)
+
+        W_i_list.append(W_i)
+        W_f_list.append(W_f)
+        W_c_list.append(W_c)
+        W_o_list.append(W_o)
+
+        U_i_list.append(U_i)
+        U_f_list.append(U_f)
+        U_c_list.append(U_c)
+        U_o_list.append(U_o)
+
+        b_i_list.append(b_i)
+        b_f_list.append(b_f)
+        b_c_list.append(b_c)
+        b_o_list.append(b_o)
+
+        print "W_i_ROWS:" + str(W_i.shape[0])
+        print "W_i_COLS:" + str(W_i.shape[1])
+        W_i_ROWS_list.append(str(W_i.shape[0]))
+        W_i_COLS_list.append(str(W_i.shape[1]))
+        print "W_i_ROWS list:" + str(W_i_ROWS_list)
+        print "W_i_COLS list:" + str(W_i_COLS_list)
+
+        W_f_ROWS_list.append(str(W_f.shape[0]))
+        W_f_COLS_list.append(str(W_f.shape[1]))
+        W_c_ROWS_list.append(str(W_c.shape[0]))
+        W_c_COLS_list.append(str(W_c.shape[1]))
+        W_o_ROWS_list.append(str(W_o.shape[0]))
+        W_o_COLS_list.append(str(W_o.shape[1]))
+        U_i_ROWS_list.append(str(U_i.shape[0]))
+        U_i_COLS_list.append(str(U_i.shape[1]))
+        U_f_ROWS_list.append(str(U_f.shape[0]))
+        U_f_COLS_list.append(str(U_f.shape[1]))
+        U_c_ROWS_list.append(str(U_c.shape[0]))
+        U_c_COLS_list.append(str(U_c.shape[1]))
+        U_o_ROWS_list.append(str(U_o.shape[0]))
+        U_o_COLS_list.append(str(U_o.shape[1]))
+        b_i_SHAPE_list.append(str(b_i.shape[0]))
+        b_f_SHAPE_list.append(str(b_f.shape[0]))
+        b_c_SHAPE_list.append(str(b_c.shape[0]))
+        b_o_SHAPE_list.append(str(b_o.shape[0]))
+
+        append_list(file, "unsigned int", 'LSTM_RECURRENT_ACTIVATION', recurrent_activation_list)
+        append_list(file, "unsigned int", 'LSTM_ACTIVATION', activation_list)
+        append_list(file, "unsigned int", 'RETURN_SEQUENCES', return_sequences_list)
+        append_list(file, "unsigned int", 'N_HIDDEN_NEURONS', num_units_list)
+        append_list(file, "unsigned int", 'W_i_ROWS', W_i_ROWS_list)
+        append_list(file, "unsigned int", 'W_i_COLS', W_i_COLS_list)
+        append_list(file, "unsigned int", 'W_f_ROWS', W_f_ROWS_list)
+        append_list(file, "unsigned int", 'W_f_COLS', W_f_COLS_list)
+        append_list(file, "unsigned int", 'W_c_ROWS', W_c_ROWS_list)
+        append_list(file, "unsigned int", 'W_c_COLS', W_c_COLS_list)
+        append_list(file, "unsigned int", 'W_o_ROWS', W_o_ROWS_list)
+        append_list(file, "unsigned int", 'W_o_COLS', W_o_COLS_list)
+        append_list(file, "unsigned int", 'U_i_ROWS', U_i_ROWS_list)
+        append_list(file, "unsigned int", 'U_i_COLS', U_i_COLS_list)
+        append_list(file, "unsigned int", 'U_f_ROWS', U_f_ROWS_list)
+        append_list(file, "unsigned int", 'U_f_COLS', U_f_COLS_list)
+        append_list(file, "unsigned int", 'U_c_ROWS', U_c_ROWS_list)
+        append_list(file, "unsigned int", 'U_c_COLS', U_c_COLS_list)
+        append_list(file, "unsigned int", 'U_o_ROWS', U_o_ROWS_list)
+        append_list(file, "unsigned int", 'U_o_COLS', U_o_COLS_list)
+        append_list(file, "unsigned int", 'b_i_SHAPE', b_i_SHAPE_list)
+        append_list(file, "unsigned int", 'b_f_SHAPE', b_f_SHAPE_list)
+        append_list(file, "unsigned int", 'b_c_SHAPE', b_c_SHAPE_list)
+        append_list(file, "unsigned int", 'b_o_SHAPE', b_o_SHAPE_list)
+
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'W_i', W_i_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'W_f', W_f_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'W_c', W_c_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'W_o', W_o_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'U_i', U_i_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'U_f', U_f_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'U_c', U_c_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'U_o', U_o_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'b_i', b_i_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'b_f', b_f_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'b_c', b_c_list)
+        append_vector_of_vectors(file, lstm_layers_count ,'float', 'b_o', b_o_list)
+
+        file.close()
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('-m', '--model', help="onnx model filename")
+    parser.add_argument("-w", "--weights", help="show initializer weights", action="store_true")
+    parser.add_argument('-l', '--lstmlayer', help="the number of LSTM layers, 1-3 layers are supported")
+    args = parser.parse_args()
+    # check if the number of LSTM layers is set and in the right range
+    lstm_layers_count = 0
+    if args.lstmlayer is not None:
+        lstm_layers_count = int(args.lstmlayer)
+    if lstm_layers_count > 3 or lstm_layers_count < 1:
+        print("1-3 LSTM layers are supported. Please set -l or --lstmlayer accordingly.")
+        sys.exit()
+    # Load the ONNX model
+    model = onnx.load(args.model)
+    # Check that the IR is well formed
+    onnx.checker.check_model(model)
+    print("================")
+    make_header(model, lstm_layers_count)
diff --git a/models/preprocessingheader.py b/models/preprocessingheader.py
new file mode 100644
index 0000000..c23c732
--- /dev/null
+++ b/models/preprocessingheader.py
@@ -0,0 +1,108 @@
+"""
+Library for data preprocessing.
+"""
+
+import numpy as np
+import pandas as pd
+
+from scipy import stats
+
+from configheader import * # Global variables
+
+def get_convoluted_data(data,
+                        segment_time_size=SEGMENT_TIME_SIZE,
+                        time_step=TIME_STEP):
+    """
+    Take as input pandas dataframe and return a tuple: convoluted data and labels.
+    Convoluted data is basically the data extracted by sliding a window of size
+    SEGMENT_TIME_SIZE with stepsize TIME_STEP. Each window is assigned with one
+    label, which is the most frequently occuring label for in given window.
+    """
+    data_convoluted = []
+    labels = []
+
+    increment = 0
+    if(len(data) == segment_time_size):
+        increment = 1
+    elif(len(data) < segment_time_size):
+        raise ValueError
+
+    # Slide a "segment_time_size" wide window with a step size of "TIME_STEP"
+    #   Increment allows the loop to run if there is just one sample, for which
+    #   len(data) == segment_time_size
+    for i in range(0, len(data) - segment_time_size + increment, time_step):
+        ax = data[COLUMN_NAMES[1]].values[i: i + segment_time_size]
+        ay = data[COLUMN_NAMES[2]].values[i: i + segment_time_size]
+        az = data[COLUMN_NAMES[3]].values[i: i + segment_time_size]
+
+        gx = data[COLUMN_NAMES[4]].values[i: i + segment_time_size]
+        gy = data[COLUMN_NAMES[5]].values[i: i + segment_time_size]
+        gz = data[COLUMN_NAMES[6]].values[i: i + segment_time_size]
+
+        mx = data[COLUMN_NAMES[7]].values[i: i + segment_time_size]
+        my = data[COLUMN_NAMES[8]].values[i: i + segment_time_size]
+        mz = data[COLUMN_NAMES[9]].values[i: i + segment_time_size]
+
+        # When 9 features used
+        # data_convoluted.append([ax, ay, az, gx, gy, gz, mx, my, mz])
+
+        # When 3 features used
+        data_convoluted.append([ax, ay, az])
+
+        # Label for a data window is the label that appears most commonly
+        label = stats.mode(data[COLUMN_NAMES[0]][i: i + segment_time_size])[0][0]
+        labels.append(label)
+
+    data_convoluted = np.asarray(data_convoluted, dtype=np.float32)
+    data_convoluted = data_convoluted.transpose(0, 2, 1)
+
+    labels = one_hot_encode(labels)
+    return data_convoluted, labels
+
+def one_hot_encode(labels):
+    """
+    Take as input a list of labels (strings) and encode them using one_hot_encoding scheme,
+    i.e. assuming there are 3 activities: A, B in the LABELS_NAMES array,
+    if the input is [A, B] then return [[1, 0, 0].[0, 1, 0]].
+    If the label does not exist then raise a NameError.
+    """
+    encoded = []
+    for i, label in enumerate(labels):
+        array = np.zeros(N_CLASSES)
+        try:
+            label_position(label)
+        except NameError:
+            raise NameError
+        array[label_position(label)] = 1
+        encoded.append(array)
+
+    return np.asarray(encoded, dtype=np.float32)
+
+def label_position(label):
+    """
+    Take as input a label (string) and if it's in LABELS_NAMES array,
+    return its position, otherwise raise NameError.
+    """
+    for i in range(N_CLASSES):
+        if(LABELS_NAMES[i] == label):
+            return i
+    raise NameError('Label does not exist')
+
+def one_hot_to_label(array):
+    """
+    Take as input a one-hot encoded array and return a label corresponding to "one".
+    i.e for [0, 0, 1] return label at index 2.
+    """
+    i = np.argmax(array)
+    return LABELS_NAMES[i]
+
+def softmax_to_one_hot(array):
+    """
+    Take an array of numbers as input and return one-hot encoded version of the
+    array, where "one" corresponds to the highest number in the array.
+    i.e. for [0, 2, 5, 3] return [0, 0, 1, 0]
+    """
+    i = np.argmax(array)
+    one_hot = np.zeros(len(array))
+    one_hot[i] = 1
+    return one_hot
diff --git a/models/readme.txt b/models/readme.txt
new file mode 100644
index 0000000..64add3a
--- /dev/null
+++ b/models/readme.txt
@@ -0,0 +1,14 @@
+===================================================
+LSTM_model.h created from the ONNX model
+===================================================
+
+The ONNX model is converted with python scripts in this folder
+to create the CPP header file (LSTM_model.h) with initialized 
+data structures, which is then compiled to establish the 
+Recurrent Neural Network (RNN) inference libarary for anomaly 
+detection.
+
+  python ./make_header.py -m ./timotor20190227raw.onnx -l 1
+
+  -m/--model: onnx model filename
+  -l/--lstmlayer: the number of LSTM layers, 1-3 layers are supported
diff --git a/models/timotor20190227raw.onnx b/models/timotor20190227raw.onnx
new file mode 100644
index 0000000..266a23d
--- /dev/null
+++ b/models/timotor20190227raw.onnx
diff --git a/motor-pdm.cpp b/motor-pdm.cpp
new file mode 100644
index 0000000..5b30193
--- /dev/null
+++ b/motor-pdm.cpp
@@ -0,0 +1,187 @@
+#include <stdio.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <string.h>
+#include <memory.h>
+#include <fcntl.h> 
+#include <sys/stat.h> 
+#include <sys/types.h> 
+#include <unistd.h>
+
+#include <iostream>
+#include <thread>
+#include <chrono>
+
+//////////////////////////////////////////////////////////////
+// Filter Code Definitions
+//////////////////////////////////////////////////////////////
+// maximum number of inputs that can be handled
+// in one function call
+#define SAMPLES 200
+// maximum length of filter than can be handled
+#define FIRFLT_LEN 1001 
+// buffer to hold all of the input samples
+#define BUFFER_LEN (FIRFLT_LEN - 1 + SAMPLES)
+
+// delay line  to hold input samples
+double insampPH1[ BUFFER_LEN ];
+double insampPH2[ BUFFER_LEN ];
+// LP filter coefficients 
+#define DRATIO 200
+double coeffs[ FIRFLT_LEN ];
+
+// Relative prediction error threshold
+#define RELATIVE_THRESHOLD 2.0
+
+// Each sample tick is 20ms long
+#define START_FIND_THRESHOLD 150
+#define STOP_FIND_THRESHOLD  1500
+#define HANGOVER_TIME        100
+
+#define max(a,b) (a > b ? a : b)
+
+int uart_get_data(int sample_count, int op_mode, float *samples_array);
+void uart_stream_parser(int mode_op);
+void lstmSetup(void);
+void runLstm(double lstm_in1, double lstm_in2, double *lstm_out1, double *lstm_out2);
+
+// FIR init
+void firFloatInit( double *insamp, int len_delay_line)
+{
+  int i;
+  for (i = 0; i < len_delay_line; i ++) insamp[i] = 0.0;
+}
+
+// the FIR filter function
+void firFloat( double *coeffs, double *insamp, int filterLength,
+               double *input,  int length,
+               double *output)
+{
+    double acc;     // accumulator for MACs
+    double *coeffp; // pointer to coefficients
+    double *inputp; // pointer to input samples
+    int n,j=0;
+    int k;
+     // put the new samples at the high end of the buffer
+    memcpy( &insamp[filterLength - 1], input,
+    length * sizeof(double) );
+    // apply the filter to each input sample
+    for ( n = 0; n < length; n=n+DRATIO ) {
+      // calculate output n
+      coeffp = coeffs;
+      inputp = &insamp[filterLength - 1 + n];
+      acc = 0;
+      for ( k = 0; k < filterLength; k++ )
+      {
+        acc += (*coeffp++) * (*inputp--);
+      }
+      output[j++] = acc;
+    }
+    // shift input samples back in time for next time
+    memmove( &insamp[0], &insamp[length], (filterLength - 1) * sizeof(double) );
+}
+//////////////////////////////////////////////////////////////
+// Test program
+//////////////////////////////////////////////////////////////
+// number of samples to read per loop
+int main( int argc, char *argv[])
+{
+  int size;
+  double floatInput1[SAMPLES];
+  double floatInput2[SAMPLES];
+  double floatOutput1[SAMPLES/DRATIO];
+  double floatOutput2[SAMPLES/DRATIO];
+  double predict_threshold_err = 0.0;
+  double predict1 = 0.0;
+  double predict2 = 0.0;
+  double mu1, mu2, sig1, sig2;
+  FILE *coeff_fid;
+  int    samples_counter = 0;
+  int    restart_hangover_timer = 0;
+  float  samples_array[3*SAMPLES];
+
+  //Pick up decimating FIR filter coefficients
+  coeff_fid = fopen( "coeff.bin", "rb" );
+  if ( coeff_fid == 0 ) {
+    printf("\n...couldn't open coeff.bin file (Decimating FIR filter coefficients)\n");
+    return -1;
+  }
+  fread(coeffs, sizeof(double), FIRFLT_LEN, coeff_fid );
+  fclose( coeff_fid );
+
+  //Get normalization parameters
+  if(argc < 5) {
+    printf("\nUsage:%s mu1 mu2 sig1 sig2\n", argv[0]);
+    return -1;
+  }
+  mu1  = atof(argv[1]);
+  mu2  = atof(argv[2]);
+  sig1 = atof(argv[3]);
+  sig2 = atof(argv[4]);
+
+  lstmSetup();
+  // initialize the filter delay lines
+  firFloatInit(insampPH1, BUFFER_LEN);
+  firFloatInit(insampPH2, BUFFER_LEN);
+
+  //Start UART reading, and parsing thread. Parsed data are logged into a buffer
+  std::thread sensorStream (uart_stream_parser, 0); 
+  // Processs loop
+  while(1)
+  {
+    // read samples from UART pipe
+    size = uart_get_data(SAMPLES, 0, samples_array);
+    if(size == SAMPLES)
+    {  // It means, at least this many SAMPLES are available 
+       for(int i = 0; i < SAMPLES; i ++) {
+         floatInput1[i] = (double)samples_array[3*i + 0];
+         floatInput2[i] = (double)samples_array[3*i + 1];
+         // position is samples_array[3*i + 2]
+       }
+       // perform the filtering
+       // One decimated sample generated
+       firFloat( coeffs, insampPH1, FIRFLT_LEN, floatInput1, SAMPLES, floatOutput1 );
+       firFloat( coeffs, insampPH2, FIRFLT_LEN, floatInput2, SAMPLES, floatOutput2 );
+       //Calculate prediction error (using previously predicted samples)
+       double predict_error  = (floatOutput1[0] - predict1) * (floatOutput1[0] - predict1);
+              predict_error += (floatOutput2[0] - predict2) * (floatOutput2[0] - predict2);
+       if(samples_counter > START_FIND_THRESHOLD)
+       {
+          if(samples_counter < STOP_FIND_THRESHOLD)
+          { // Find error threshold
+            predict_threshold_err = max(predict_error, predict_threshold_err);
+            std::cout << "Find error threshold:" << samples_counter << std::endl;
+            restart_hangover_timer = 0;
+          } else {
+            std::cout << "Look for anomaly:" << samples_counter << " err:" << predict_error << " thre:" << (RELATIVE_THRESHOLD * predict_threshold_err);
+            // Anomaly detection mode!
+            // ...implemented hangover
+            if(predict_error > RELATIVE_THRESHOLD * predict_threshold_err)
+            {
+               restart_hangover_timer = samples_counter;
+            }
+            std::cout << std::endl;
+            if((samples_counter - restart_hangover_timer) < HANGOVER_TIME)
+            {
+              std::cout << " [AD] ";
+            }
+          }
+       }
+       //Normalize
+       double lstm_out1, lstm_in1 = (floatOutput1[0] - mu1) / sig1;
+       double lstm_out2, lstm_in2 = (floatOutput2[0] - mu2) / sig2;
+       //Run LSTM network
+       runLstm(lstm_in1, lstm_in2, &lstm_out1, &lstm_out2);
+       //Denormalize
+       predict1 = lstm_out1 * sig1 + mu1;
+       predict2 = lstm_out2 * sig2 + mu2;
+
+       samples_counter ++; // Increamented at decimated rate!!!
+    } else {
+      std::this_thread::sleep_for(std::chrono::milliseconds(100));
+      std::cout << "sleep" << std::endl;
+    }
+  } 
+ 
+  return 0;
+}
diff --git a/onnx_model.cpp b/onnx_model.cpp
new file mode 100644
index 0000000..5713708
--- /dev/null
+++ b/onnx_model.cpp
@@ -0,0 +1,389 @@
+/*
+ * Copyright (c) 2016 Robert W. Rose
+ *
+ * Edited by Tomasz Bartkowiak (July 2018)
+ * MIT License, see LICENSE file.
+ *
+ * Modified by Texas Instruments (2019) for ONNX import support
+ * MIT License.
+ */
+
+#include "onnx_model.h"
+
+#ifdef DBG_START
+#include "start_LSTM_model.h"
+#else
+#include "LSTM_model.h"
+#endif
+
+#include <iostream>
+#include <cmath>
+#include <limits>
+#include <stdio.h>
+
+bool OnnxLayerActivation::LoadActivation(ActivationType activation_) {
+    switch (activation_) {
+    case kLinear:
+        activation_type_ = kLinear;
+        break;
+    case kRelu:
+        activation_type_ = kRelu;
+        break;
+    case kSoftPlus:
+        activation_type_ = kSoftPlus;
+        break;
+    case kHardSigmoid:
+        activation_type_ = kHardSigmoid;
+        break;
+    case kSigmoid:
+        activation_type_ = kSigmoid;
+        break;
+    case kTanh:
+        activation_type_ = kTanh;
+        break;
+    default:
+        KASSERT(false, "Unsupported activation type %d", activation_);
+    }
+
+    return true;
+}
+
+bool OnnxLayerActivation::Apply(Tensor* in, Tensor* out) {
+    KASSERT(in, "Invalid input");
+    KASSERT(out, "Invalid output");
+
+    *out = *in;
+#ifdef DBG_VERBOSE
+    std::cout << "Activation Apply(" << activation_type_ << ")" << std::endl;
+#endif
+    switch (activation_type_) {
+    case kLinear:
+        break;
+    case kRelu:
+        for (size_t i = 0; i < out->data_.size(); i++) {
+            if (out->data_[i] < 0.0) {
+                out->data_[i] = 0.0;
+            }
+        }
+        break;
+    case kSoftPlus:
+        for (size_t i = 0; i < out->data_.size(); i++) {
+            out->data_[i] = std::log(1.0 + std::exp(out->data_[i]));
+        }
+        break;
+    case kHardSigmoid:
+        for (size_t i = 0; i < out->data_.size(); i++) {
+            float x = (out->data_[i] * 0.2) + 0.5;
+
+            if (x <= 0) {
+                out->data_[i] = 0.0;
+            } else if (x >= 1) {
+                out->data_[i] = 1.0;
+            } else {
+                out->data_[i] = x;
+            }
+        }
+        break;
+    case kSigmoid:
+        for (size_t i = 0; i < out->data_.size(); i++) {
+            float& x = out->data_[i];
+
+            if (x >= 0) {
+                out->data_[i] = 1.0 / (1.0 + std::exp(-x));
+            } else {
+                float z = std::exp(x);
+                out->data_[i] = z / (1.0 + z);
+            }
+        }
+        break;
+    case kTanh:
+        for (size_t i = 0; i < out->data_.size(); i++) {
+            out->data_[i] = std::tanh(out->data_[i]);
+        }
+        break;
+    default:
+        break;
+    }
+
+    return true;
+}
+
+bool OnnxLayerDense::LoadLayer(unsigned int dummy_int) {
+    unsigned int weights_rows = DENSE_WEIGHTS_ROWS;
+    unsigned int weights_cols = DENSE_WEIGHTS_COLS;
+    unsigned int biases_shape = DENSE_BIASES_SHAPE;
+
+    weights_.Resize(weights_rows, weights_cols);
+    weights_.data_ = DENSE_WEIGHTS;
+    biases_.Resize(biases_shape);
+    biases_.data_ = DENSE_BIASES;
+    activation_.LoadActivation(ActivationType::kLinear);
+    return true;
+}
+
+bool OnnxLayerDense::Apply(Tensor* in, Tensor* out) {
+    KASSERT(in, "Invalid input");
+    KASSERT(out, "Invalid output");
+    KASSERT(in->dims_.size() <= 2, "Invalid input dimensions");
+
+    if (in->dims_.size() == 2) {
+        KASSERT(in->dims_[1] == weights_.dims_[0], "Dimension mismatch %d %d",
+                in->dims_[1], weights_.dims_[0]);
+    }
+
+#ifdef DBG_VERBOSE
+    std::cout << "DenseLayer Weights Size:" << weights_.dims_[0] << " " << weights_.dims_[1] << std::endl;
+    std::cout << "DenseLayer Input Size  :" << in->dims_[0] << " " << in->dims_[1] << std::endl << std::flush;
+#endif
+
+#if 1
+    Tensor tmp(weights_.dims_[1]);
+
+    for (int i = 0; i < weights_.dims_[0]; i++) {
+        for (int j = 0; j < weights_.dims_[1]; j++) {
+            tmp(j) += (*in)(i)*weights_(i, j);
+        }
+    }
+
+    for (int i = 0; i < biases_.dims_[0]; i++) {
+        tmp(i) += biases_(i);
+    }
+#else
+    std::cout << "BiasesDIM DenseLayer:" << biases_.dims_[0] << std::endl;
+    Tensor tmp = in->Dot(weights_) + biases_;
+    tmp.Print();
+#endif
+    KASSERT(activation_.Apply(&tmp, out), "Failed to apply activation");
+
+#ifdef DBG_VERBOSE
+    std::cout << "DenseLayer TMP Size:" << tmp.dims_[0] << std::endl;
+    std::cout << "DenseLayer Output Dimensions:" << out->dims_.size() << std::endl;
+    if(out->dims_.size() == 1)
+      std::cout << "DenseLayer Output Size :" << out->dims_[0] << std::endl;
+    if(out->dims_.size() == 2)
+      std::cout << "DenseLayer Output Size :" << out->dims_[0] << " " << out->dims_[1] << std::endl;
+#endif
+    return true;
+}
+
+bool OnnxLayerLSTM::LoadLayer(unsigned int i) {
+    unsigned int wi_rows = W_i_ROWS[i];
+    unsigned int wi_cols = W_i_COLS[i];
+    unsigned int ui_rows = U_i_ROWS[i];
+    unsigned int ui_cols = U_i_COLS[i];
+    unsigned int bi_shape = b_i_SHAPE[i];
+    unsigned int wf_rows = W_f_ROWS[i];
+    unsigned int wf_cols = W_f_COLS[i];
+    unsigned int uf_rows = U_f_ROWS[i];
+    unsigned int uf_cols = U_f_COLS[i];
+    unsigned int bf_shape = b_f_SHAPE[i];
+    unsigned int wc_rows = W_c_ROWS[i];
+    unsigned int wc_cols = W_c_COLS[i];
+    unsigned int uc_rows = U_c_ROWS[i];
+    unsigned int uc_cols = U_c_COLS[i];
+    unsigned int bc_shape = b_c_SHAPE[i];
+    unsigned int wo_rows = W_o_ROWS[i];
+    unsigned int wo_cols = W_o_COLS[i];
+    unsigned int uo_rows = U_o_ROWS[i];
+    unsigned int uo_cols = U_o_COLS[i];
+    unsigned int bo_shape = b_o_SHAPE[i];
+
+    // Load Input Weights and Biases
+    Wi_.Resize(wi_rows, wi_cols);
+    Ui_.Resize(ui_rows, ui_cols);
+    bi_.Resize(1, bi_shape);
+    Wi_.data_ = W_i[i];
+    Ui_.data_ = U_i[i];
+    bi_.data_ = b_i[i];
+
+    // Load Forget Weights and Biases
+    Wf_.Resize(wf_rows, wf_cols);
+    Uf_.Resize(uf_rows, uf_cols);
+    bf_.Resize(1, bf_shape);
+    Wf_.data_ = W_f[i];
+    Uf_.data_ = U_f[i];
+    bf_.data_ = b_f[i];
+
+    // Load State Weights and Biases
+    Wc_.Resize(wc_rows, wc_cols);
+    Uc_.Resize(uc_rows, uc_cols);
+    bc_.Resize(1, bc_shape);
+    Wc_.data_ = W_c[i];
+    Uc_.data_ = U_c[i];
+    bc_.data_ = b_c[i];
+
+    // Load Output Weights and Biases
+    Wo_.Resize(wo_rows, wo_cols);
+    Uo_.Resize(uo_rows, uo_cols);
+    bo_.Resize(1, bo_shape);
+    Wo_.data_ = W_o[i];
+    Uo_.data_ = U_o[i];
+    bo_.data_ = b_o[i];
+
+    innerActivation_.LoadActivation(ActivationType::kSigmoid);
+    activation_.LoadActivation(ActivationType::kTanh);
+    return_sequences_ = (bool)RETURN_SEQUENCES[i];
+
+#ifdef DO_LSTM_RESET
+    Reset();
+#endif
+
+    return true;
+}
+
+#ifdef DO_LSTM_RESET
+bool OnnxLayerLSTM::Reset() {
+    // Assume bo always keeps the output shape and we will always receive one
+    // single sample.
+#ifdef DBG_VERBOSE
+    std::cout << "Reset LSTM enter" << std::endl;
+#endif
+    int outputDim = bo_.dims_[1];
+    ht_1.Resize(1, outputDim);
+    ct_1.Resize(1, outputDim);
+
+    ht_1.Fill(0.0f);
+    ct_1.Fill(0.0f);
+#ifdef DBG_VERBOSE
+    std::cout << "Reset LSTM end:" << outputDim << std::endl;
+#endif
+    return true;
+}
+#endif
+
+bool OnnxLayerLSTM::Apply(Tensor* in, Tensor* out) {
+    // Assume bo always keeps the output shape and we will always receive one
+    // single sample.
+    int outputDim = bo_.dims_[1];
+#ifndef DO_LSTM_RESET
+    Tensor ht_1 = Tensor(1, outputDim);
+    Tensor ct_1 = Tensor(1, outputDim);
+
+    ht_1.Fill(0.0f);
+    ct_1.Fill(0.0f);
+#endif
+
+    int steps = in->dims_[0];
+
+    Tensor outputs, lastOutput;
+
+    if (return_sequences_) {
+        outputs.dims_ = {steps, outputDim};
+        outputs.data_.reserve(steps * outputDim);
+    }
+
+    for (int s = 0; s < steps; s++) {
+        Tensor x = in->Select(s);
+
+        KASSERT(Step(&x, &lastOutput, &ht_1, &ct_1), "Failed to execute step");
+
+        if (return_sequences_) {
+            outputs.data_.insert(outputs.data_.end(), lastOutput.data_.begin(),
+                                 lastOutput.data_.end());
+        }
+    }
+
+    if (return_sequences_) {
+        *out = outputs;
+        std::cout << "Return dim(" << outputs.dims_[0] << "," << outputs.dims_[1] << ")" << std::endl;
+    } else {
+        *out = lastOutput;
+    }
+
+    return true;
+}
+
+bool OnnxLayerLSTM::Step(Tensor* x, Tensor* out, Tensor* ht_1, Tensor* ct_1) {
+    Tensor xi = x->Dot(Wi_) + bi_;
+    Tensor xf = x->Dot(Wf_) + bf_;
+    Tensor xc = x->Dot(Wc_) + bc_;
+    Tensor xo = x->Dot(Wo_) + bo_;
+#ifdef DBG_VERBOSE
+    printf("\nXi[%d,%d]", xi.dims_[0], xi.dims_[1]);
+    printf("\nXf[%d,%d]", xf.dims_[0], xf.dims_[1]);
+    printf("\nXc[%d,%d]", xc.dims_[0], xc.dims_[1]);
+    printf("\nXo[%d,%d]", xo.dims_[0], xo.dims_[1]);
+#endif
+
+    Tensor i_ = xi + ht_1->Dot(Ui_);
+    Tensor f_ = xf + ht_1->Dot(Uf_);
+    Tensor c_ = xc + ht_1->Dot(Uc_);
+    Tensor o_ = xo + ht_1->Dot(Uo_);
+
+#ifdef DBG_VERBOSE
+    printf("\ni[%d,%d]", i_.dims_[0], i_.dims_[1]);
+    printf("\nf[%d,%d]", f_.dims_[0], f_.dims_[1]);
+    printf("\nc[%d,%d]", c_.dims_[0], c_.dims_[1]);
+    printf("\no[%d,%d]", o_.dims_[0], o_.dims_[1]);
+    std::cout << "\nLSTM step" << std::endl;
+#endif
+
+    Tensor i, f, cc, o;
+
+    KASSERT(innerActivation_.Apply(&i_, &i),
+            "Failed to apply inner activation on i");
+    KASSERT(innerActivation_.Apply(&f_, &f),
+            "Failed to apply inner activation on f");
+    KASSERT(activation_.Apply(&c_, &cc), "Failed to apply activation on c_");
+    KASSERT(innerActivation_.Apply(&o_, &o),
+            "Failed to apply inner activation on o");
+
+    *ct_1 = f.Multiply(*ct_1) + i.Multiply(cc);
+
+    KASSERT(activation_.Apply(ct_1, &cc), "Failed to apply activation on c");
+    *out = *ht_1 = o.Multiply(cc);
+
+    return true;
+}
+
+bool OnnxModel::LoadModel() {
+    unsigned int num_layers = NUM_LAYERS;
+
+    LayerType layer_types[num_layers] = {LayerType::kLSTM, LayerType::kDense};
+    for (unsigned int i = 0; i < num_layers; i++) {
+        LayerType layer_type = layer_types[i];
+        OnnxLayer* layer = NULL;
+
+        switch (layer_type) {
+        case kDense:
+            layer = new OnnxLayerDense();
+            break;
+        case kActivation:
+            layer = new OnnxLayerActivation();
+            break;
+        case kLSTM:
+            layer = new OnnxLayerLSTM();
+            break;
+        default:
+            break;
+        }
+
+        bool result = layer->LoadLayer(i);
+        if (!result) {
+            printf("Failed to load layer %d", i);
+            delete layer;
+            return false;
+        }
+        layers_.push_back(layer);
+    }
+
+    return true;
+}
+
+bool OnnxModel::Apply(Tensor* in, Tensor* out) {
+    Tensor temp_in, temp_out;
+
+    for (unsigned int i = 0; i < layers_.size(); i++) {
+        if (i == 0) {
+            temp_in = *in;
+        }
+        KASSERT(layers_[i]->Apply(&temp_in, &temp_out),
+                "Failed to apply layer %d", i);
+
+        temp_in = temp_out;
+    }
+
+    *out = temp_out;
+
+    return true;
+}
diff --git a/onnx_model.h b/onnx_model.h
new file mode 100644
index 0000000..3a3d165
--- /dev/null
+++ b/onnx_model.h
@@ -0,0 +1,397 @@
+/*
+ * Copyright (c) 2016 Robert W. Rose
+ *
+ * MIT License, see LICENSE file.
+ *
+ * Modified by Texas Instruments (2019) for ONNX import support
+ * MIT License.
+ */
+
+#ifndef ONNX_MODEL_H_
+#define ONNX_MODEL_H_
+
+#define DO_LSTM_RESET
+
+#include <algorithm>
+#include <numeric>
+#include <string>
+#include <vector>
+
+#define KASSERT(x, ...)                                                        \
+    if (!(x)) {                                                                \
+        printf("KASSERT: %s(%d): ", __FILE__, __LINE__);                       \
+        printf(__VA_ARGS__);                                                   \
+        printf("\n");                                                          \
+        return false;                                                          \
+    }
+
+#define KASSERT_EQ(x, y, eps)                                                  \
+    if (fabs(x - y) > eps) {                                                   \
+        printf("KASSERT: Expected %f, got %f\n", y, x);                        \
+        return false;                                                          \
+    }
+
+#ifdef DEBUG
+#define KDEBUG(x, ...)                                                         \
+    if (!(x)) {                                                                \
+        printf("%s(%d): ", __FILE__, __LINE__);                                \
+        printf(__VA_ARGS__);                                                   \
+        printf("\n");                                                          \
+        exit(-1);                                                              \
+    }
+#else
+#define KDEBUG(x, ...) ;
+#endif
+
+class Tensor {
+  public:
+    Tensor() {}
+
+    Tensor(int i) { Resize(i); }
+
+    Tensor(int i, int j) { Resize(i, j); }
+
+    Tensor(int i, int j, int k) { Resize(i, j, k); }
+
+    Tensor(int i, int j, int k, int l) { Resize(i, j, k, l); }
+
+    void Resize(int i) {
+        dims_ = {i};
+        data_.resize(i);
+    }
+
+    void Resize(int i, int j) {
+        dims_ = {i, j};
+        data_.resize(i * j);
+    }
+
+    void Resize(int i, int j, int k) {
+        dims_ = {i, j, k};
+        data_.resize(i * j * k);
+    }
+
+    void Resize(int i, int j, int k, int l) {
+        dims_ = {i, j, k, l};
+        data_.resize(i * j * k * l);
+    }
+
+    inline void Flatten() {
+        KDEBUG(dims_.size() > 0, "Invalid tensor");
+
+        int elements = dims_[0];
+        for (unsigned int i = 1; i < dims_.size(); i++) {
+            elements *= dims_[i];
+        }
+        dims_ = {elements};
+    }
+
+    inline float& operator()(int i) {
+        KDEBUG(dims_.size() == 1, "Invalid indexing for tensor");
+        KDEBUG(i < dims_[0] && i >= 0, "Invalid i: %d (max %d)", i, dims_[0]);
+
+        return data_[i];
+    }
+
+    inline float& operator()(int i, int j) {
+        KDEBUG(dims_.size() == 2, "Invalid indexing for tensor");
+        KDEBUG(i < dims_[0] && i >= 0, "Invalid i: %d (max %d)", i, dims_[0]);
+        KDEBUG(j < dims_[1] && j >= 0, "Invalid j: %d (max %d)", j, dims_[1]);
+
+        return data_[dims_[1] * i + j];
+    }
+
+    inline float operator()(int i, int j) const {
+        KDEBUG(dims_.size() == 2, "Invalid indexing for tensor");
+        KDEBUG(i < dims_[0] && i >= 0, "Invalid i: %d (max %d)", i, dims_[0]);
+        KDEBUG(j < dims_[1] && j >= 0, "Invalid j: %d (max %d)", j, dims_[1]);
+
+        return data_[dims_[1] * i + j];
+    }
+
+    inline float& operator()(int i, int j, int k) {
+        KDEBUG(dims_.size() == 3, "Invalid indexing for tensor");
+        KDEBUG(i < dims_[0] && i >= 0, "Invalid i: %d (max %d)", i, dims_[0]);
+        KDEBUG(j < dims_[1] && j >= 0, "Invalid j: %d (max %d)", j, dims_[1]);
+        KDEBUG(k < dims_[2] && k >= 0, "Invalid k: %d (max %d)", k, dims_[2]);
+
+        return data_[dims_[2] * (dims_[1] * i + j) + k];
+    }
+
+    inline float& operator()(int i, int j, int k, int l) {
+        KDEBUG(dims_.size() == 4, "Invalid indexing for tensor");
+        KDEBUG(i < dims_[0] && i >= 0, "Invalid i: %d (max %d)", i, dims_[0]);
+        KDEBUG(j < dims_[1] && j >= 0, "Invalid j: %d (max %d)", j, dims_[1]);
+        KDEBUG(k < dims_[2] && k >= 0, "Invalid k: %d (max %d)", k, dims_[2]);
+        KDEBUG(l < dims_[3] && l >= 0, "Invalid l: %d (max %d)", l, dims_[3]);
+
+        return data_[dims_[3] * (dims_[2] * (dims_[1] * i + j) + k) + l];
+    }
+
+    inline void Fill(float value) {
+        std::fill(data_.begin(), data_.end(), value);
+    }
+
+    Tensor Unpack(int row) const {
+        KASSERT(dims_.size() >= 2, "Invalid tensor");
+        std::vector<int> pack_dims =
+            std::vector<int>(dims_.begin() + 1, dims_.end());
+        int pack_size = std::accumulate(pack_dims.begin(), pack_dims.end(), 0);
+
+        std::vector<float>::const_iterator first =
+            data_.begin() + (row * pack_size);
+        std::vector<float>::const_iterator last =
+            data_.begin() + (row + 1) * pack_size;
+
+        Tensor x = Tensor();
+        x.dims_ = pack_dims;
+        x.data_ = std::vector<float>(first, last);
+
+        return x;
+    }
+
+    Tensor Select(int row) const {
+        Tensor x = Unpack(row);
+        x.dims_.insert(x.dims_.begin(), 1);
+
+        return x;
+    }
+
+    Tensor operator+(const Tensor& other) {
+        KASSERT(dims_ == other.dims_,
+                "Cannot add tensors with different dimensions");
+
+        Tensor result;
+        result.dims_ = dims_;
+        result.data_.reserve(data_.size());
+
+        std::transform(data_.begin(), data_.end(), other.data_.begin(),
+                       std::back_inserter(result.data_),
+                       [](float x, float y) { return x + y; });
+
+        return result;
+    }
+
+    Tensor Multiply(const Tensor& other) {
+        KASSERT(dims_ == other.dims_,
+                "Cannot multiply elements with different dimensions");
+
+        Tensor result;
+        result.dims_ = dims_;
+        result.data_.reserve(data_.size());
+
+        std::transform(data_.begin(), data_.end(), other.data_.begin(),
+                       std::back_inserter(result.data_),
+                       [](float x, float y) { return x * y; });
+
+        return result;
+    }
+
+    Tensor Dot(const Tensor& other) {
+        KDEBUG(dims_.size() == 2, "Invalid tensor dimensions");
+        KDEBUG(other.dims_.size() == 2, "Invalid tensor dimensions");
+        KASSERT(dims_[1] == other.dims_[0],
+                "Cannot multiply with different inner dimensions");
+
+        Tensor tmp(dims_[0], other.dims_[1]);
+
+        for (int i = 0; i < dims_[0]; i++) {
+            for (int j = 0; j < other.dims_[1]; j++) {
+                for (int k = 0; k < dims_[1]; k++) {
+                    tmp(i, j) += (*this)(i, k) * other(k, j);
+                }
+            }
+        }
+
+        return tmp;
+    }
+
+    void Print() {
+        if (dims_.size() == 1) {
+            printf("[ ");
+            for (int i = 0; i < dims_[0]; i++) {
+                printf("%f ", (*this)(i));
+            }
+            printf("]\n");
+        } else if (dims_.size() == 2) {
+            printf("[\n");
+            for (int i = 0; i < dims_[0]; i++) {
+                printf(" [ ");
+                for (int j = 0; j < dims_[1]; j++) {
+                    printf("%f ", (*this)(i, j));
+                }
+                printf("]\n");
+            }
+            printf("]\n");
+        } else if (dims_.size() == 3) {
+            printf("[\n");
+            for (int i = 0; i < dims_[0]; i++) {
+                printf(" [\n");
+                for (int j = 0; j < dims_[1]; j++) {
+                    printf("  [ ");
+                    for (int k = 0; k < dims_[2]; k++) {
+                        printf("%f ", (*this)(i, j, k));
+                    }
+                    printf("  ]\n");
+                }
+                printf(" ]\n");
+            }
+            printf("]\n");
+        } else if (dims_.size() == 4) {
+            printf("[\n");
+            for (int i = 0; i < dims_[0]; i++) {
+                printf(" [\n");
+                for (int j = 0; j < dims_[1]; j++) {
+                    printf("  [\n");
+                    for (int k = 0; k < dims_[2]; k++) {
+                        printf("   [");
+                        for (int l = 0; l < dims_[3]; l++) {
+                            printf("%f ", (*this)(i, j, k, l));
+                        }
+                        printf("]\n");
+                    }
+                    printf("  ]\n");
+                }
+                printf(" ]\n");
+            }
+            printf("]\n");
+        }
+    }
+
+    void PrintShape() {
+        printf("(");
+        for (unsigned int i = 0; i < dims_.size(); i++) {
+            printf("%d ", dims_[i]);
+        }
+        printf(")\n");
+    }
+
+    std::vector<int> dims_;
+    std::vector<float> data_;
+};
+
+enum ActivationType {
+    kLinear = 1,
+    kRelu = 2,
+    kSoftPlus = 3,
+    kSigmoid = 4,
+    kTanh = 5,
+    kHardSigmoid = 6
+};
+
+enum LayerType {
+    kDense = 1,
+    // kConvolution2d = 2,
+    // kFlatten = 3,
+    // kElu = 4,
+    kActivation = 2,
+    // kMaxPooling2D = 6,
+    kLSTM = 3,
+    // kEmbedding = 8
+};
+
+class OnnxLayer {
+  public:
+    OnnxLayer() {}
+
+    virtual ~OnnxLayer() {}
+
+    virtual bool LoadLayer(unsigned int i) = 0;
+
+    virtual bool Apply(Tensor* in, Tensor* out) = 0;
+
+    virtual bool Reset() = 0;
+};
+
+class OnnxLayerActivation : public OnnxLayer {
+  public:
+    OnnxLayerActivation() : activation_type_(ActivationType::kLinear) {}
+
+    virtual ~OnnxLayerActivation() {}
+
+    virtual bool LoadLayer(unsigned int i) { return true; }
+
+    bool LoadActivation(ActivationType activation_);
+
+    virtual bool Apply(Tensor* in, Tensor* out);
+
+    virtual bool Reset() { return true; }
+
+  private:
+    ActivationType activation_type_;
+};
+
+class OnnxLayerDense : public OnnxLayer {
+  public:
+    OnnxLayerDense() {}
+
+    virtual ~OnnxLayerDense() {}
+
+    virtual bool LoadLayer(unsigned int dummy_int);
+
+    virtual bool Apply(Tensor* in, Tensor* out);
+
+    virtual bool Reset() { return true; }
+
+  private:
+    Tensor weights_;
+    Tensor biases_;
+
+    OnnxLayerActivation activation_;
+};
+
+class OnnxLayerLSTM : public OnnxLayer {
+  public:
+    OnnxLayerLSTM() : return_sequences_(false) {}
+
+    virtual ~OnnxLayerLSTM() {}
+
+    virtual bool LoadLayer(unsigned int i);
+
+    virtual bool Apply(Tensor* in, Tensor* out);
+#ifdef DO_LSTM_RESET
+    virtual bool Reset();
+#else
+    virtual bool Reset() { return true; }
+#endif
+  private:
+    bool Step(Tensor* x, Tensor* out, Tensor* ht_1, Tensor* ct_1);
+
+#ifdef DO_LSTM_RESET
+    Tensor ht_1;
+    Tensor ct_1;
+#endif
+    Tensor Wi_;
+    Tensor Ui_;
+    Tensor bi_;
+    Tensor Wf_;
+    Tensor Uf_;
+    Tensor bf_;
+    Tensor Wc_;
+    Tensor Uc_;
+    Tensor bc_;
+    Tensor Wo_;
+    Tensor Uo_;
+    Tensor bo_;
+    OnnxLayerActivation innerActivation_;
+    OnnxLayerActivation activation_;
+    bool return_sequences_;
+};
+
+class OnnxModel {
+  public:
+    OnnxModel() {}
+    virtual ~OnnxModel() {
+        for (unsigned int i = 0; i < layers_.size(); i++) {
+            delete layers_[i];
+        }
+    }
+
+    virtual bool LoadModel();
+
+    virtual bool Apply(Tensor* in, Tensor* out);
+
+  private:
+    std::vector<OnnxLayer*> layers_;
+};
+
+#endif // ONNX_MODEL_H_
diff --git a/psensors.cpp b/psensors.cpp
new file mode 100644
index 0000000..8b333c9
--- /dev/null
+++ b/psensors.cpp
@@ -0,0 +1,144 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <memory.h>
+#include <fcntl.h> 
+#include <sys/stat.h> 
+#include <sys/types.h> 
+#include <unistd.h>
+
+#include <iostream>
+#include <thread>
+#include <chrono>
+
+//#define VERBOSE
+
+#define MAX_BUFFER (3*64*16384)
+static int rd_data_cnt = 0;
+static int wr_data_cnt = 0;
+static float data_ready_to_use[MAX_BUFFER];
+
+int uart_get_data(int samples_count, int overlap_mode, float *samples_array)
+{
+   int rd_idx;
+#ifdef VERBOSE
+   printf ("\nwr=%d rd=%d cnt=%d ", wr_data_cnt, rd_data_cnt, samples_count);
+   printf ("\noverlap_mode=%d ", overlap_mode);
+#endif
+   if(wr_data_cnt >= (rd_data_cnt + samples_count))
+   {
+      for (int i = 0; i < samples_count; i ++)
+      {
+         rd_idx = rd_data_cnt % MAX_BUFFER;
+         samples_array[3*i + 0] = data_ready_to_use[rd_idx + 0];
+         samples_array[3*i + 1] = data_ready_to_use[rd_idx + 1];
+         samples_array[3*i + 2] = data_ready_to_use[rd_idx + 2];
+         rd_data_cnt += 3;
+      }
+      return samples_count;
+   } else return 0;
+}
+
+void uart_stream_parser(int mode_op) {
+const char *log_filename = mode_op & 0x01 ? "parser.txt" : NULL;
+FILE *fout;
+int wr_idx = 0;
+#ifdef VERBOSE
+int prev_sync = 0;
+#endif
+int prev_cnt  = 0;
+unsigned int val1, val_phases, val_position;
+short int_phase1, int_phase2;
+float phase1, phase2, position;
+unsigned char all_zeros[256];
+unsigned char line_in[256];
+int sync_i = (1 << 31);
+int cnt, err_flag = 0;
+int fd;
+// FIFO file path 
+const char * myfifo = "/tmp/myfifo"; 
+
+  memset(all_zeros, 0, 256);
+  memset(line_in, 0, 256);
+
+  // Creating the named file(FIFO) 
+  mkfifo(myfifo, 0666); 
+
+  // Open FIFO for Read only 
+  fd = open(myfifo, O_RDONLY); 
+
+  if(log_filename)
+  {
+     fout = fopen(log_filename, "wt");
+  }
+
+  for (int i = 0; ; i ++)
+  { // Shift delay line left for one byte
+    for(int j = 1; j < 256; j ++) line_in[j - 1] = line_in[j];
+    if(read(fd, &line_in[255], 1) != 1) {
+      line_in[255] = 0;
+      if(wr_data_cnt <= rd_data_cnt) { // Continue processing until all buffered samples are processed
+        // No unprocessed bytes left, sleep for some time as there is nothing on input
+        std::this_thread::sleep_for(std::chrono::milliseconds(20));
+        continue;
+      }
+    }
+
+    // Writting too fast, do write side throttling
+    if((mode_op & 0x02) && (wr_data_cnt > (rd_data_cnt + MAX_BUFFER - 16)))
+       std::this_thread::sleep_for(std::chrono::milliseconds(100));
+
+    if(line_in[0] == 0xce && 
+       line_in[1] == 0xfa &&
+       line_in[2] == 0xbe &&
+       line_in[3] == 0xba
+      )
+    {
+       cnt = *(unsigned int *)&line_in[1*4];
+       sync_i = i + 2*4;
+#ifdef VERBOSE
+       printf ("\n%08x, %8d:%8d ... %8d[%8d]\n", *(unsigned int *)&line_in[0], i, cnt, 
+                                                 sync_i - prev_sync, cnt - prev_cnt);
+       prev_sync = sync_i;
+#endif
+       if((i >= 4*1024) && (cnt - prev_cnt) != 1023) err_flag |= 1;
+       prev_cnt  = cnt;
+    }
+    if((i - sync_i) >= 0)
+    {
+#ifdef VERBOSE
+      if(((i - sync_i) % (2*4)) == 0) printf ("\n[%8d, %8d] ... ", i, sync_i); 
+#endif
+      if(((i - sync_i) % 4)  == 0) 
+      {
+         val1 = *(unsigned int *)&line_in[0];
+#ifdef VERBOSE
+         printf ("%08x  ", val1);
+#endif
+         if(((i - sync_i) % 8) == 0) {
+           val_phases = val1;
+         } else {
+           val_position = (int)val1;
+           int_phase1 = (short)(val_phases >> 16);
+           int_phase2 = (short)((val_phases << 16) >> 16);
+           phase1     = (float)int_phase1 / 32767.0;
+           phase2     = (float)int_phase2 / 32767.0;
+           position   = (float)val_position / 33554432.0f;
+           if(log_filename) fprintf(fout, "%8.5f %8.5f %8.5f\n", phase1, phase2, position);
+#ifdef VERBOSE
+           printf("... %8.5f %8.5f %8.5f", phase1, phase2, position);
+#endif
+           wr_idx = wr_data_cnt % MAX_BUFFER;
+           data_ready_to_use[wr_idx + 0] = phase1;
+           data_ready_to_use[wr_idx + 1] = phase2;
+           data_ready_to_use[wr_idx + 2] = position;
+           wr_data_cnt += 3;
+         }
+      }
+    }
+  }
+  printf("\nEnd of stream!!!\n"); fflush(stdout);
+  if(log_filename) fclose(fout);
+  close(fd);
+}
+/** nothing past this point **/
diff --git a/readme.txt b/readme.txt
new file mode 100644
index 0000000..162f32e
--- /dev/null
+++ b/readme.txt
@@ -0,0 +1,17 @@
+==============================
+Native compilation on target
+==============================
+Copy the folder pdm-anomaly-detection to target, and then run "./mk.sh"
+After the compilation is completed successfully, motpdm binary will be created.
+
+
+==============================
+Run motpdm on target
+==============================
+Terminal 1: ./run.sh > log.txt
+Terminal 2: cat ./logs/normal45-270-v100-with-friction2-iter10-15.log > /tmp/myfifo
+
+After the cat command in Terminal 2 is done, stop the process (CTRL-C) in Terminal 1.
+
+Then, check reports in log.txt, with time units being in 20ms ticks (decimated sample
+rate of 50Hz). AD (Anomaly detection) is reported along with the sample counter.
diff --git a/run.sh b/run.sh
new file mode 100755
index 0000000..f040753
--- /dev/null
+++ b/run.sh
@@ -0,0 +1 @@
+./motpdm -0.0001841035315 0.004191935886 0.01918238488 0.01933241767