Skip to content

2
2026_Article_RESILINK_ML
Commit 9ceb46a7 by Joaquín Irazábal González
Options

Refactor hysteretic curves processing and prediction scripts

- Added a new function `load_tfdmap` to load TFDMap data from files, ensuring proper error handling for missing or malformed data. - Updated the hysteretic curves processing script to include TFDMap data in the point-level DataFrame, maintaining alignment with case folders. - Enhanced the prediction script to handle multiple target columns dynamically, improving flexibility for model training and evaluation. - Modified caching mechanism for window sizes to use a tuple key for better organization and retrieval. - Improved logging and error messages for better debugging and user feedback. - Updated the run script to ensure consistent execution of prediction tasks with specified parameters.
parent f2ad8555
Showing with 436 additions and 664 deletions
Code/data/hysteretic_curves/2W/H30_B29/merged_dataset_points.csv
This source diff could not be displayed because it is too large. You can view the blob instead.
Code/data/hysteretic_curves/2W/H30_B34/merged_dataset_points.csv
This source diff could not be displayed because it is too large. You can view the blob instead.
Code/data/hysteretic_curves/3W/H45_B29/merged_dataset_points.csv
This source diff could not be displayed because it is too large. You can view the blob instead.
Code/data/hysteretic_curves/3W/H45_B34/merged_dataset_points.csv
This source diff could not be displayed because it is too large. You can view the blob instead.
Code/data/hysteretic_curves/5W/H60_B34/merged_dataset_points.csv
This source diff could not be displayed because it is too large. You can view the blob instead.
Code/reports/hysteretic_curves/2W/H30_B29/case_12_with_rnn_preds.csv
Case,tw1,tw2,Displ,CumDispl,LoadDir,CycleNum,MaxAmpl,Force,Force_RNN Case,tw1,tw2,Displ,CumDispl,LoadDir,CycleNum,MaxAmpl,Force,TFDMapW1,TFDMapW2,TFDMapF,Force_RNN
12,12.76,15.0,0.0,0.0,1.0,1,0.049600873,0.0, 12,12.76,15.0,0.0,0.0,1.0,1,0.049600873,0.0,0.0,0.0,0.0,
12,12.76,15.0,0.0009600229,0.0009600229,1.0,1,0.049600873,0.0, 12,12.76,15.0,0.0047999951,0.0047999951,1.0,1,0.049600873,-289811.98,0.0,0.0,0.056928,
12,12.76,15.0,0.0019200301,0.0019200301,1.0,1,0.049600873,-73125.12, 12,12.76,15.0,0.0095999885,0.0095999885,1.0,1,0.049600873,-377965.75,0.470046,0.639734,0.683278,
12,12.76,15.0,0.0028800159,0.0028800159,1.0,1,0.049600873,-151487.12, 12,12.76,15.0,0.009600078,0.014208046,-1.0,2,0.049600873,-184305.38,0.785147,0.974071,0.975768,
12,12.76,15.0,0.0038399855,0.0038399855,1.0,1,0.049600873,-225485.12, 12,12.76,15.0,0.0048000575,0.0190080665,-1.0,2,0.049600873,59143.247,0.785147,0.974071,0.98076,
12,12.76,15.0,0.0047999951,0.0047999951,1.0,1,0.049600873,-289811.98, 12,12.76,15.0,7.8867329e-08,0.023808045132671,-1.0,2,0.049600873,287977.49,0.851529,1.020403,1.248425,
12,12.76,15.0,0.0057599741,0.0057599741,1.0,1,0.049600873,-329528.94, 12,12.76,15.0,-0.0047999289,0.0286080529,-1.0,2,0.049600873,361809.08,1.356749,1.702603,1.88383,
12,12.76,15.0,0.0067199625,0.0067199625,1.0,1,0.049600873,-347578.5, 12,12.76,15.0,-0.0095999443,0.0334080682999999,-1.0,2,0.049600873,399997.11,2.014539,2.400686,2.469553,
12,12.76,15.0,0.0076800027,0.0076800027,1.0,1,0.049600873,-359537.43, 12,12.76,15.0,-0.0095999076,0.0380160584,1.0,3,0.049600873,207052.03,2.325355,2.709506,2.736589,
12,12.76,15.0,0.0086399859,0.0086399859,1.0,1,0.049600873,-369359.6, 12,12.76,15.0,-0.0047998948,0.0428160712,1.0,3,0.049600873,-36588.629,2.325355,2.709506,2.74145,
12,12.76,15.0,0.0095999885,0.0095999885,1.0,1,0.049600873,-377965.75, 12,12.76,15.0,6.8131205e-08,0.047616034131205,1.0,3,0.049600873,-279940.85,2.378317,2.724726,2.9574,
12,12.76,15.0,0.010559995,0.010559995,1.0,1,0.049600873,-385727.06, 12,12.76,15.0,0.0048000654,0.0524160313999999,1.0,3,0.049600873,-369655.17,2.845332,3.246218,3.488382,
12,12.76,15.0,0.011519997,0.011519997,1.0,1,0.049600873,-392382.95, 12,12.76,15.0,0.0096000768,0.0572160428,1.0,3,0.049600873,-405778.62,3.465901,3.951151,4.104137,
12,12.76,15.0,0.011520085,0.012288039,-1.0,2,0.049600873,-345240.24, 12,12.76,15.0,0.0096001757,0.0618241442999999,-1.0,4,0.049600873,-211157.82,3.790219,4.291442,4.401733,
12,12.76,15.0,0.010560089,0.013248035,-1.0,2,0.049600873,-263238.67, 12,12.76,15.0,0.0048002256,0.0666240944,-1.0,4,0.049600873,31540.194,3.790219,4.291442,4.404056,
12,12.76,15.0,0.009600078,0.014208046,-1.0,2,0.049600873,-184305.38, 12,12.76,15.0,1.7269626e-07,0.07142414730374,-1.0,4,0.049600873,276756.37,3.836265,4.299633,4.602282,
12,12.76,15.0,0.0086400498,0.0151680742,-1.0,2,0.049600873,-106730.64, 12,12.76,15.0,-0.0047998111,0.0762241311,-1.0,4,0.049600873,370649.0,4.287116,4.798409,5.112079,
12,12.76,15.0,0.0076800899,0.0161280340999999,-1.0,2,0.049600873,-32994.613, 12,12.76,15.0,-0.0095998312,0.0810241512,-1.0,4,0.049600873,407371.07,4.902065,5.503762,5.721379,
12,12.76,15.0,0.0067200305,0.0170880935,-1.0,2,0.049600873,0.0, 12,12.76,15.0,-0.0095998283,0.0856321357000001,1.0,5,0.049600873,214496.42,5.22577,5.846686,6.017521,
12,12.76,15.0,0.0057600554,0.0180480686,-1.0,2,0.049600873,1274.1645,11563.1328125 12,12.76,15.0,-0.0047998582,0.0904321058000001,1.0,5,0.049600873,-27814.934,5.22577,5.846686,6.022228,-6528.94921875
12,12.76,15.0,0.0048000575,0.0190080665,-1.0,2,0.049600873,59143.247,58353.359375 12,12.76,15.0,1.5436796e-07,0.0952321183679601,1.0,5,0.049600873,-272754.91,5.266926,5.851645,6.230035,-265175.625
12,12.76,15.0,0.003840103,0.019968021,-1.0,2,0.049600873,117899.36,119181.9609375 12,12.76,15.0,0.010000188,0.1052321520000001,1.0,5,0.049600873,-418303.32,6.382042,7.098776,7.409964,-394390.5625
12,12.76,15.0,0.002880074,0.02092805,-1.0,2,0.049600873,172293.01,172919.15625 12,12.76,15.0,0.020000155,0.1152321190000001,1.0,5,0.049600873,-462593.31,7.618197,8.354361,8.545721,-447724.25
12,12.76,15.0,0.0019201034,0.0218880206,-1.0,2,0.049600873,220117.37,222401.484375 12,12.76,15.0,0.020000272,0.1248320880000001,-1.0,6,0.049600873,-71734.652,8.172086,8.889643,9.060265,-81217.7890625
12,12.76,15.0,0.00096008975,0.02284803425,-1.0,2,0.049600873,258703.81,258541.453125 12,12.76,15.0,0.01000028,0.1348320800000001,-1.0,6,0.049600873,311565.26,8.442799,9.117646,9.561961,302632.9375
12,12.76,15.0,7.8867329e-08,0.023808045132671,-1.0,2,0.049600873,287977.49,284805.9375 12,12.76,15.0,1.5892522e-07,0.1448322010747801,-1.0,6,0.049600873,395245.42,9.423494,10.19456,10.662497,402579.4375
12,12.76,15.0,-0.00095990783,0.02476803183,-1.0,2,0.049600873,309649.37,305397.3125 12,12.76,15.0,-0.0099997658,0.1548321258000001,-1.0,6,0.049600873,460211.56,10.709636,11.52057,11.770478,461116.375
12,12.76,15.0,-0.001919891,0.025728015,-1.0,2,0.049600873,325858.74,322626.84375 12,12.76,15.0,-0.019999834,0.164832194,-1.0,6,0.049600873,499908.52,11.864922,12.610947,12.772742,494937.75
12,12.76,15.0,-0.0028799375,0.0266880615,-1.0,2,0.049600873,339047.44,337545.09375 12,12.76,15.0,-0.019999946,0.174431986,1.0,7,0.049600873,107920.86,12.414463,13.135973,13.282014,119000.328125
12,12.76,15.0,-0.003839951,0.027648075,-1.0,2,0.049600873,351058.48,350782.40625 12,12.76,15.0,-0.0099997785,0.1844321535,1.0,7,0.049600873,-304913.4,12.628146,13.292991,13.721678,-288156.8125
12,12.76,15.0,-0.0047999289,0.0286080529,-1.0,2,0.049600873,361809.08,361918.53125 12,12.76,15.0,1.9164559e-07,0.1944321236455901,1.0,7,0.049600873,-405669.06,13.520816,14.255532,14.780805,-399593.875
12,12.76,15.0,-0.0057599624,0.0295680864,-1.0,2,0.049600873,371157.27,370323.1875 12,12.76,15.0,0.010000109,0.2044320410000001,1.0,7,0.049600873,-475301.62,14.789293,15.571822,15.907708,-480244.8125
12,12.76,15.0,-0.00671997,0.0305280939999999,-1.0,2,0.049600873,379405.7,378727.84375 12,12.76,15.0,0.020000143,0.2144320750000003,1.0,7,0.049600873,-505330.13,16.000783,16.726672,16.976107,-516234.1875
12,12.76,15.0,-0.0076799603,0.0314880843,-1.0,2,0.049600873,386813.52,389233.625 12,12.76,15.0,0.020000201,0.2240321490000003,-1.0,8,0.049600873,-107426.53,16.580634,17.273795,17.497072,-114822.34375
12,12.76,15.0,-0.0086399481,0.0324480720999999,-1.0,2,0.049600873,393522.49,388603.28125 12,12.76,15.0,0.010000213,0.2340321370000003,-1.0,8,0.049600873,307870.27,16.788,17.420432,17.905947,297863.3125
12,12.76,15.0,-0.0095999443,0.0334080682999999,-1.0,2,0.049600873,399997.11,400369.78125 12,12.76,15.0,2.177377e-07,0.2440321322623003,-1.0,8,0.049600873,412038.32,17.695412,18.391117,18.950234,407782.6875
12,12.76,15.0,-0.010559948,0.034368072,-1.0,2,0.049600873,406233.63,404362.0 12,12.76,15.0,-0.0099997517,0.2540321017000004,-1.0,8,0.049600873,480636.77,18.987376,19.729206,20.075531,473691.0
12,12.76,15.0,-0.011519962,0.035328086,-1.0,2,0.049600873,411882.43,399949.5625 12,12.76,15.0,-0.01999973,0.2640320800000004,-1.0,8,0.049600873,510098.54,20.21402,20.897078,21.142058,515317.3125
12,12.76,15.0,-0.011519962,0.036096004,1.0,3,0.049600873,365636.16,348261.0 12,12.76,15.0,-0.019999908,0.2736325080000004,1.0,9,0.049600873,112806.33,20.795068,21.454233,21.669493,145341.96875
12,12.76,15.0,-0.010559962,0.037056004,1.0,3,0.049600873,284305.32,277662.0 12,12.76,15.0,-0.0099999645,0.2836324515000005,1.0,9,0.049600873,-303699.37,21.001226,21.598277,22.097359,-289240.8125
12,12.76,15.0,-0.0095999076,0.0380160584,1.0,3,0.049600873,207052.03,197502.734375 12,12.76,15.0,6.0256808e-08,0.2936324762568083,1.0,9,0.049600873,-408093.21,21.905376,22.564824,23.155952,-408049.1875
12,12.76,15.0,-0.008639969,0.038975997,1.0,3,0.049600873,129930.93,117658.6171875 12,12.76,15.0,0.015199983,0.3088323990000005,1.0,9,0.049600873,-505072.49,23.840373,24.511011,24.850612,-496288.25
12,12.76,15.0,-0.0076799712,0.0399359948,1.0,3,0.049600873,57076.753,52995.3984375 12,12.76,15.0,0.030399915,0.3240323310000006,1.0,9,0.049600873,-545131.55,25.5655,26.153381,26.429387,-537047.375
12,12.76,15.0,-0.0067199087,0.0408960573,1.0,3,0.049600873,0.0,11182.2978515625 12,12.76,15.0,0.030399993,0.3386247190000006,-1.0,10,0.049600873,25226.638,26.478181,27.057659,27.230124,809.343994140625
12,12.76,15.0,-0.005759939,0.041856027,1.0,3,0.049600873,0.0,-16553.025390625 12,12.76,15.0,0.015200128,0.3538245840000006,-1.0,10,0.049600873,384300.17,27.365417,27.891973,28.371927,375262.1875
12,12.76,15.0,-0.0047998948,0.0428160712,1.0,3,0.049600873,-36588.629,-46862.2734375 12,12.76,15.0,-2.9466151e-08,0.3690247414661515,-1.0,10,0.049600873,467305.05,28.873296,29.372548,29.929423,471956.5625
12,12.76,15.0,-0.0038399758,0.0437759901999999,1.0,3,0.049600873,-94568.829,-93928.2734375 12,12.76,15.0,-0.015199802,0.3842245140000003,-1.0,10,0.049600873,541019.29,30.756317,31.244145,31.547355,540466.5
12,12.76,15.0,-0.0028799434,0.0447360226,1.0,3,0.049600873,-150514.62,-150239.375 12,12.76,15.0,-0.030399956,0.3994246680000002,-1.0,10,0.049600873,578944.97,32.470902,32.858169,33.113327,577323.125
12,12.76,15.0,-0.0019199088,0.0456960572,1.0,3,0.049600873,-201342.41,-199511.59375 12,12.76,15.0,-0.030400175,0.4140165590000003,1.0,11,0.049600873,7043.9335,33.402315,33.820487,33.965346,14372.271484375
12,12.76,15.0,-0.00095997256,0.04665599344,1.0,3,0.049600873,-244843.12,-242165.140625 12,12.76,15.0,-0.015200034,0.4292167000000004,1.0,11,0.049600873,-386023.9,34.203135,34.560337,35.077852,-369458.1875
12,12.76,15.0,6.8131205e-08,0.047616034131205,1.0,3,0.049600873,-279940.85,-280196.1875 12,12.76,15.0,-1.9112045e-07,0.4444165428795504,1.0,11,0.049600873,-472109.3,35.660223,35.98214,36.650857,-455962.8125
12,12.76,15.0,0.00096010865,0.04857607465,1.0,3,0.049600873,-306909.65,-310452.90625 12,12.76,15.0,0.01519978,0.4596165140000005,1.0,11,0.049600873,-550963.29,37.523316,37.830723,38.273459,-540516.1875
12,12.76,15.0,0.0019200577,0.0495360237,1.0,3,0.049600873,-327340.74,-333145.4375 12,12.76,15.0,0.030399941,0.4748166750000005,1.0,11,0.049600873,-583317.66,39.25601,39.464577,39.880519,-587345.75
12,12.76,15.0,0.0028801068,0.0504960727999999,1.0,3,0.049600873,-343487.44,-349954.71875 12,12.76,15.0,0.030400173,0.4894088410000004,-1.0,12,0.049600873,-5716.2657,40.207903,40.467612,40.734218,-15323.044921875
12,12.76,15.0,0.0038400806,0.0514560465999999,1.0,3,0.049600873,-357557.72,-362561.6875 12,12.76,15.0,0.015200207,0.5046088070000004,-1.0,12,0.049600873,391467.38,41.003578,41.198965,41.803809,383934.3125
12,12.76,15.0,0.0048000654,0.0524160313999999,1.0,3,0.049600873,-369655.17,-372647.25 12,12.76,15.0,5.9955722e-08,0.5198089540442782,-1.0,12,0.049600873,479590.99,42.490777,42.64704,43.363202,483663.9375
12,12.76,15.0,0.0057600514,0.0533760173999999,1.0,3,0.049600873,-379743.8,-380421.53125 12,12.76,15.0,-0.015199751,0.5350087650000003,-1.0,12,0.049600873,556303.83,44.366392,44.513131,44.978406,559111.625
12,12.76,15.0,0.0067200322,0.0543359982,1.0,3,0.049600873,-387883.56,-386514.90625 12,12.76,15.0,-0.030399936,0.5502089500000004,-1.0,12,0.049600873,588198.81,46.110903,46.151415,46.581383,609410.0
12,12.76,15.0,0.0076800881,0.0552960541,1.0,3,0.049600873,-394858.79,-391767.8125 12,12.76,15.0,-0.030400406,0.5648015820000004,1.0,13,0.049600873,9480.0309,47.06955,47.165274,47.460719,21845.220703125
12,12.76,15.0,0.0086400472,0.0562560132,1.0,3,0.049600873,-400523.79,-397020.71875 12,12.76,15.0,-0.015200498,0.5800014900000005,1.0,13,0.049600873,-388477.74,47.863234,47.893506,48.565213,-387452.875
12,12.76,15.0,0.0096000768,0.0572160428,1.0,3,0.049600873,-405778.62,-402063.5 12,12.76,15.0,-3.0057512e-07,0.5952016874248803,1.0,13,0.049600873,-475186.6,49.337284,49.328373,50.146634,-478076.75
12,12.76,15.0,0.010560094,0.05817606,1.0,3,0.049600873,-410655.44,-406896.15625 12,12.76,15.0,0.019999723,0.6152017110000002,1.0,13,0.049600873,-574569.66,51.776768,51.709318,52.27778,-574337.5625
12,12.76,15.0,0.011520096,0.059136062,1.0,3,0.049600873,-414897.5,-407106.28125 12,12.76,15.0,0.039999655,0.635201643,1.0,13,0.049600873,-639410.14,54.142161,54.083376,54.418228,-622034.375
12,12.76,15.0,0.011520182,0.0599041379999999,-1.0,4,0.049600873,-367996.08,-382312.59375 12,12.76,15.0,0.040000117,0.6544023989999999,-1.0,14,0.049600873,75457.377,55.643614,55.470939,55.476663,76832.03125
12,12.76,15.0,0.010560187,0.0608641329999999,-1.0,4,0.049600873,-287543.45,-308351.71875 12,12.76,15.0,0.020000311,0.6744022050000001,-1.0,14,0.049600873,432094.72,57.112977,56.857027,57.189769,432281.5
12,12.76,15.0,0.0096001757,0.0618241442999999,-1.0,4,0.049600873,-211157.82,-219472.609375 12,12.76,15.0,2.6998025e-07,0.6944022460197499,-1.0,14,0.049600873,526580.21,59.039439,58.688025,59.185845,526591.0625
12,12.76,15.0,0.0086402332,0.0627840867999999,-1.0,4,0.049600873,-134861.59,-145721.875 12,12.76,15.0,-0.019999617,0.7144021329999997,-1.0,14,0.049600873,601259.65,61.46112,61.082516,61.370082,608976.4375
12,12.76,15.0,0.0076801853,0.0637441347,-1.0,4,0.049600873,-62499.333,-69922.5078125 12,12.76,15.0,-0.039999781,0.7344022969999996,-1.0,14,0.049600873,667991.17,63.857757,63.486548,63.572054,664044.5625
12,12.76,15.0,0.006720206,0.064704114,-1.0,4,0.049600873,0.0,-21228.107421875 12,12.76,15.0,-0.03999998,0.7536022179999996,1.0,15,0.049600873,-49058.703,65.372264,64.96707,64.707603,-34144.3046875
12,12.76,15.0,0.0057602204,0.0656640996,-1.0,4,0.049600873,0.0,8916.9833984375 12,12.76,15.0,-0.019999727,0.7736024709999997,1.0,15,0.049600873,-432177.32,66.725609,66.255224,66.459216,-419973.375
12,12.76,15.0,0.0048002256,0.0666240944,-1.0,4,0.049600873,31540.194,43553.3046875 12,12.76,15.0,5.4958764e-08,0.7936022529587639,1.0,15,0.049600873,-523867.13,68.570203,67.975814,68.471332,-509730.0625
12,12.76,15.0,0.0038401912,0.0675841288,-1.0,4,0.049600873,88295.394,99063.34375 12,12.76,15.0,0.020000335,0.813602533,1.0,15,0.049600873,-606542.93,70.958061,70.33707,70.676546,-609026.125
12,12.76,15.0,0.002880183,0.068544137,-1.0,4,0.049600873,144226.08,149911.4375 12,12.76,15.0,0.040000145,0.833602343,1.0,15,0.049600873,-668313.7,73.261157,72.756927,73.014903,-681004.9375
12,12.76,15.0,0.0019201723,0.0695041477,-1.0,4,0.049600873,195479.48,201600.0 12,12.76,15.0,0.040000821,0.8528031230000002,-1.0,16,0.049600873,55310.017,74.779705,74.321542,74.18843,34501.12890625
12,12.76,15.0,0.00096022418,0.07046409582,-1.0,4,0.049600873,240004.19,246354.71875 12,12.76,15.0,0.020000836,0.8728031080000002,-1.0,16,0.049600873,434149.53,76.140579,75.612265,75.907919,443121.6875
12,12.76,15.0,1.7269626e-07,0.07142414730374,-1.0,4,0.049600873,276756.37,279763.15625 12,12.76,15.0,5.7932109e-07,0.8928033646789101,-1.0,16,0.049600873,531866.08,78.054646,77.40104,77.907096,540900.125
12,12.76,15.0,-0.00095983223,0.07238415223,-1.0,4,0.049600873,305075.32,304556.84375 12,12.76,15.0,-0.01999925,0.9128031940000002,-1.0,16,0.049600873,628778.21,80.658909,80.040507,80.145302,628922.375
12,12.76,15.0,-0.0019197623,0.0733440823,-1.0,4,0.049600873,326677.63,324307.78125 12,12.76,15.0,-0.039999118,0.9328030620000002,-1.0,16,0.049600873,731629.97,83.439583,83.256788,82.557774,703069.1875
12,12.76,15.0,-0.0028797837,0.0743041037,-1.0,4,0.049600873,343607.91,340696.8125 12,12.76,15.0,-0.039999705,0.9520033870000002,1.0,17,0.049600873,83.169636,85.26927,85.301306,83.775157,-6610.2509765625
12,12.76,15.0,-0.0038398009,0.0752641209,-1.0,4,0.049600873,358212.51,354564.46875 12,12.76,15.0,-0.019999661,0.9720034310000004,1.0,17,0.049600873,-434395.61,86.721778,86.774096,85.592634,-435800.0625
12,12.76,15.0,-0.0047998111,0.0762241311,-1.0,4,0.049600873,370649.0,365910.75 12,12.76,15.0,3.5415353e-07,0.9920034461535306,1.0,17,0.049600873,-542005.43,88.818848,88.788557,87.670866,-530543.25
12,12.76,15.0,-0.005759811,0.077184131,-1.0,4,0.049600873,380924.28,374315.40625
12,12.76,15.0,-0.0067198051,0.0781441251,-1.0,4,0.049600873,389242.0,383560.5
12,12.76,15.0,-0.007679804,0.079104124,-1.0,4,0.049600873,396329.19,390914.5625
12,12.76,15.0,-0.0086398152,0.0800641352,-1.0,4,0.049600873,402067.93,394486.53125
12,12.76,15.0,-0.0095998312,0.0810241512,-1.0,4,0.049600873,407371.07,405622.6875
12,12.76,15.0,-0.01055976,0.08198408,-1.0,4,0.049600873,412278.88,410035.125
12,12.76,15.0,-0.011519778,0.082944098,-1.0,4,0.049600873,416528.01,407303.625
12,12.76,15.0,-0.011519868,0.083712096,1.0,5,0.049600873,369854.31,357296.0
12,12.76,15.0,-0.010559876,0.084672088,1.0,5,0.049600873,290181.63,287117.21875
12,12.76,15.0,-0.0095998283,0.0856321357000001,1.0,5,0.049600873,214496.42,206747.84375
12,12.76,15.0,-0.0086398972,0.0865920668000001,1.0,5,0.049600873,138884.31,125958.203125
12,12.76,15.0,-0.0076799073,0.0875520567,1.0,5,0.049600873,66887.207,58799.8515625
12,12.76,15.0,-0.006719854,0.08851211,1.0,5,0.049600873,629.76786,14675.4775390625
12,12.76,15.0,-0.0057598955,0.0894720685000001,1.0,5,0.049600873,0.0,-13611.400390625
12,12.76,15.0,-0.0047998582,0.0904321058000001,1.0,5,0.049600873,-27814.934,-42975.1171875
12,12.76,15.0,-0.0038398607,0.0913921033000001,1.0,5,0.049600873,-83140.925,-89095.6015625
12,12.76,15.0,-0.0028798336,0.0923521304000001,1.0,5,0.049600873,-138967.58,-147507.859375
12,12.76,15.0,-0.0019198869,0.0933120771000001,1.0,5,0.049600873,-190231.74,-200562.171875
12,12.76,15.0,-0.00095986541,0.0942720985900001,1.0,5,0.049600873,-235261.69,-246577.578125
12,12.76,15.0,1.5436796e-07,0.0952321183679601,1.0,5,0.049600873,-272754.91,-287550.25
12,12.76,15.0,0.0020001588,0.0972321228000001,1.0,5,0.049600873,-333523.8,-325161.03125
12,12.76,15.0,0.0040001755,0.0992321395000001,1.0,5,0.049600873,-367082.15,-356888.53125
12,12.76,15.0,0.00600004,0.1012320040000001,1.0,5,0.049600873,-390968.09,-382102.46875
12,12.76,15.0,0.0080000526,0.1032320166000001,1.0,5,0.049600873,-406756.71,-401853.375
12,12.76,15.0,0.010000188,0.1052321520000001,1.0,5,0.049600873,-418303.32,-416561.5
12,12.76,15.0,0.01200008,0.1072320440000001,1.0,5,0.049600873,-427609.39,-428328.0
12,12.76,15.0,0.014000051,0.1092320150000001,1.0,5,0.049600873,-435661.82,-437573.125
12,12.76,15.0,0.016000068,0.1112320320000001,1.0,5,0.049600873,-444516.41,-446818.21875
12,12.76,15.0,0.018000104,0.1132320680000001,1.0,5,0.049600873,-453624.01,-456903.78125
12,12.76,15.0,0.020000155,0.1152321190000001,1.0,5,0.049600873,-462593.31,-469931.0
12,12.76,15.0,0.022000036,0.1172320000000001,1.0,5,0.049600873,-471308.99,-485059.34375
12,12.76,15.0,0.024000109,0.1192320730000001,1.0,5,0.049600873,-479746.49,-477915.40625
12,12.76,15.0,0.024000292,0.1208320680000001,-1.0,6,0.049600873,-383953.62,-392608.28125
12,12.76,15.0,0.022000317,0.1228320430000001,-1.0,6,0.049600873,-221509.43,-235231.328125
12,12.76,15.0,0.020000272,0.1248320880000001,-1.0,6,0.049600873,-71734.652,-89200.65625
12,12.76,15.0,0.018000141,0.1268322190000001,-1.0,6,0.049600873,24901.57,27998.1484375
12,12.76,15.0,0.016000145,0.1288322150000001,-1.0,6,0.049600873,141879.29,138775.28125
12,12.76,15.0,0.014000254,0.1308321060000001,-1.0,6,0.049600873,226958.62,219985.15625
12,12.76,15.0,0.012000307,0.1328320530000001,-1.0,6,0.049600873,279260.42,273459.6875
12,12.76,15.0,0.01000028,0.1348320800000001,-1.0,6,0.049600873,311565.26,310440.09375
12,12.76,15.0,0.008000236,0.1368321240000001,-1.0,6,0.049600873,332888.17,330401.125
12,12.76,15.0,0.0060002174,0.1388321426000001,-1.0,6,0.049600873,348990.11,342587.875
12,12.76,15.0,0.0040002289,0.1408321311000001,-1.0,6,0.049600873,362473.33,356455.53125
12,12.76,15.0,0.0020001437,0.1428322163000001,-1.0,6,0.049600873,377256.14,377257.03125
12,12.76,15.0,1.5892522e-07,0.1448322010747801,-1.0,6,0.049600873,395245.42,396167.46875
12,12.76,15.0,-0.0019997257,0.1468320857000001,-1.0,6,0.049600873,411228.24,413186.875
12,12.76,15.0,-0.0039998562,0.1488322162000001,-1.0,6,0.049600873,427008.47,427684.875
12,12.76,15.0,-0.0059998537,0.1508322137000001,-1.0,6,0.049600873,441662.44,440081.71875
12,12.76,15.0,-0.0079997082,0.1528320682000001,-1.0,6,0.049600873,452039.9,451007.75
12,12.76,15.0,-0.0099997658,0.1548321258000001,-1.0,6,0.049600873,460211.56,460462.96875
12,12.76,15.0,-0.011999851,0.1568322110000001,-1.0,6,0.049600873,468299.52,468027.15625
12,12.76,15.0,-0.013999782,0.1588321420000001,-1.0,6,0.049600873,476152.86,479373.4375
12,12.76,15.0,-0.01599973,0.16083209,-1.0,6,0.049600873,484129.34,487778.0625
12,12.76,15.0,-0.017999695,0.162832055,-1.0,6,0.049600873,492079.0,491560.15625
12,12.76,15.0,-0.019999834,0.164832194,-1.0,6,0.049600873,499908.52,499124.34375
12,12.76,15.0,-0.02199976,0.16683212,-1.0,6,0.049600873,507543.45,506268.28125
12,12.76,15.0,-0.023999815,0.168832175,-1.0,6,0.049600873,514961.8,510890.84375
12,12.76,15.0,-0.023999862,0.17043207,1.0,7,0.049600873,415746.91,407303.625
12,12.76,15.0,-0.02199991,0.172432022,1.0,7,0.049600873,257821.69,251397.5
12,12.76,15.0,-0.019999946,0.174431986,1.0,7,0.049600873,107920.86,98643.109375
12,12.76,15.0,-0.01799995,0.1764319820000001,1.0,7,0.049600873,15794.489,-19520.912109375
12,12.76,15.0,-0.015999824,0.1784321080000001,1.0,7,0.049600873,-110553.53,-112838.7109375
12,12.76,15.0,-0.01399989,0.180432042,1.0,7,0.049600873,-199710.95,-205394.84375
12,12.76,15.0,-0.011999892,0.18243204,1.0,7,0.049600873,-263295.27,-270110.59375
12,12.76,15.0,-0.0099997785,0.1844321535,1.0,7,0.049600873,-304913.4,-311083.25
12,12.76,15.0,-0.0079998688,0.1864320632,1.0,7,0.049600873,-333057.67,-335876.9375
12,12.76,15.0,-0.0059999042,0.1884320278,1.0,7,0.049600873,-353736.13,-354367.15625
12,12.76,15.0,-0.003999881,0.1904320510000001,1.0,7,0.049600873,-370058.81,-370546.09375
12,12.76,15.0,-0.001999789,0.1924321430000001,1.0,7,0.049600873,-386512.44,-387565.5
12,12.76,15.0,1.9164559e-07,0.1944321236455901,1.0,7,0.049600873,-405669.06,-406686.0625
12,12.76,15.0,0.0020000653,0.1964319973000001,1.0,7,0.049600873,-423187.56,-426647.09375
12,12.76,15.0,0.0040001921,0.1984321241000001,1.0,7,0.049600873,-439664.08,-444296.84375
12,12.76,15.0,0.006000214,0.2004321460000001,1.0,7,0.049600873,-455151.16,-459004.96875
12,12.76,15.0,0.0080001322,0.2024320642000001,1.0,7,0.049600873,-467333.92,-469931.0
12,12.76,15.0,0.010000109,0.2044320410000001,1.0,7,0.049600873,-475301.62,-477915.40625
12,12.76,15.0,0.012000136,0.2064320680000002,1.0,7,0.049600873,-482804.56,-483378.40625
12,12.76,15.0,0.014000142,0.2084320740000002,1.0,7,0.049600873,-488885.66,-487160.5
12,12.76,15.0,0.016000178,0.2104321100000003,1.0,7,0.049600873,-494666.52,-490522.375
12,12.76,15.0,0.018000108,0.2124320400000003,1.0,7,0.049600873,-500132.27,-494304.46875
12,12.76,15.0,0.020000143,0.2144320750000003,1.0,7,0.049600873,-505330.13,-501028.15625
12,12.76,15.0,0.022000139,0.2164320710000003,1.0,7,0.049600873,-510283.31,-511954.21875
12,12.76,15.0,0.024000134,0.2184320660000003,1.0,7,0.049600873,-515047.29,-515316.09375
12,12.76,15.0,0.024000294,0.2200320560000003,-1.0,8,0.049600873,-415249.31,-434631.5
12,12.76,15.0,0.022000291,0.2220320590000003,-1.0,8,0.049600873,-257005.01,-275153.40625
12,12.76,15.0,0.020000201,0.2240321490000003,-1.0,8,0.049600873,-107426.53,-110632.4921875
12,12.76,15.0,0.018000211,0.2260321390000003,-1.0,8,0.049600873,-15435.138,1779.603515625
12,12.76,15.0,0.016000204,0.2280321460000003,-1.0,8,0.049600873,110483.45,102215.0859375
12,12.76,15.0,0.014000169,0.2300321810000003,-1.0,8,0.049600873,200116.13,195821.796875
12,12.76,15.0,0.012000264,0.2320320860000003,-1.0,8,0.049600873,264950.49,256860.515625
12,12.76,15.0,0.010000213,0.2340321370000003,-1.0,8,0.049600873,307870.27,306447.90625
12,12.76,15.0,0.0080002459,0.2360321041000003,-1.0,8,0.049600873,337046.97,336074.28125
12,12.76,15.0,0.0060002378,0.2380321122000003,-1.0,8,0.049600873,358331.18,352463.3125
12,12.76,15.0,0.0040002383,0.2400321117000003,-1.0,8,0.049600873,375104.65,371583.875
12,12.76,15.0,0.00200016,0.2420321900000003,-1.0,8,0.049600873,392067.9,393646.0625
12,12.76,15.0,2.177377e-07,0.2440321322623003,-1.0,8,0.049600873,412038.32,413607.09375
12,12.76,15.0,-0.0019997543,0.2460321043000003,-1.0,8,0.049600873,429657.96,431046.71875
12,12.76,15.0,-0.0039997454,0.2480320954000003,-1.0,8,0.049600873,445535.87,445754.84375
12,12.76,15.0,-0.0059997252,0.2500320752000003,-1.0,8,0.049600873,460998.98,457941.59375
12,12.76,15.0,-0.0079998472,0.2520321972000003,-1.0,8,0.049600873,472748.42,468447.40625
12,12.76,15.0,-0.0099997517,0.2540321017000004,-1.0,8,0.049600873,480636.77,477272.28125
12,12.76,15.0,-0.011999816,0.2560321660000003,-1.0,8,0.049600873,487977.68,483995.96875
12,12.76,15.0,-0.01399985,0.2580322000000004,-1.0,8,0.049600873,493990.28,491139.9375
12,12.76,15.0,-0.015999688,0.2600320380000004,-1.0,8,0.049600873,499658.17,499124.34375
12,12.76,15.0,-0.017999709,0.2620320590000004,-1.0,8,0.049600873,505012.41,502065.96875
12,12.76,15.0,-0.01999973,0.2640320800000004,-1.0,8,0.049600873,510098.54,507108.75
12,12.76,15.0,-0.021999756,0.2660321060000004,-1.0,8,0.049600873,514936.01,514672.9375
12,12.76,15.0,-0.023999776,0.2680321260000004,-1.0,8,0.049600873,519541.74,517614.5625
12,12.76,15.0,-0.023999992,0.2696324240000003,1.0,9,0.049600873,419324.13,416758.84375
12,12.76,15.0,-0.022000034,0.2716323820000004,1.0,9,0.049600873,262049.86,263164.0
12,12.76,15.0,-0.019999908,0.2736325080000004,1.0,9,0.049600873,112806.33,109411.5625
12,12.76,15.0,-0.017999953,0.2756324630000004,1.0,9,0.049600873,44919.279,-12560.818359375
12,12.76,15.0,-0.016000018,0.2776323980000004,1.0,9,0.049600873,-103369.37,-104644.1875
12,12.76,15.0,-0.013999889,0.2796325270000004,1.0,9,0.049600873,-194825.76,-196149.734375
12,12.76,15.0,-0.011999894,0.2816325220000005,1.0,9,0.049600873,-260028.52,-264857.71875
12,12.76,15.0,-0.0099999645,0.2836324515000005,1.0,9,0.049600873,-303699.37,-309402.3125
12,12.76,15.0,-0.0079999077,0.2856325083000005,1.0,9,0.049600873,-333371.56,-336507.28125
12,12.76,15.0,-0.0059999948,0.2876324212000004,1.0,9,0.049600873,-354868.72,-357729.0
12,12.76,15.0,-0.0040000244,0.2896323916000003,1.0,9,0.049600873,-372051.94,-376219.21875
12,12.76,15.0,-0.0020000108,0.2916324052000004,1.0,9,0.049600873,-388819.4,-395760.03125
12,12.76,15.0,6.0256808e-08,0.2936324762568083,1.0,9,0.049600873,-408093.21,-417401.96875
12,12.76,15.0,0.0030399591,0.2966723751000004,1.0,9,0.049600873,-442824.43,-442615.90625
12,12.76,15.0,0.0060801387,0.2997125547000004,1.0,9,0.049600873,-467756.56,-466989.375
12,12.76,15.0,0.0091200559,0.3027524719000004,1.0,9,0.049600873,-484343.79,-486740.28125
12,12.76,15.0,0.012159904,0.3057923200000004,1.0,9,0.049600873,-496174.66,-501028.15625
12,12.76,15.0,0.015199983,0.3088323990000005,1.0,9,0.049600873,-505072.49,-511954.21875
12,12.76,15.0,0.018240008,0.3118724240000005,1.0,9,0.049600873,-513386.61,-519938.59375
12,12.76,15.0,0.021280116,0.3149125320000005,1.0,9,0.049600873,-521062.27,-526242.0625
12,12.76,15.0,0.024320113,0.3179525290000005,1.0,9,0.049600873,-528232.05,-532125.3125
12,12.76,15.0,0.027360034,0.3209924500000006,1.0,9,0.049600873,-535953.09,-540109.75
12,12.76,15.0,0.030399915,0.3240323310000006,1.0,9,0.049600873,-545131.55,-553136.9375
12,12.76,15.0,0.033439916,0.3270723320000006,1.0,9,0.049600873,-555855.48,-574148.5625
12,12.76,15.0,0.036480018,0.3301124340000006,1.0,9,0.049600873,-577304.65,-583393.625
12,12.76,15.0,0.036480092,0.3325446200000006,-1.0,10,0.049600873,-430421.12,-445137.28125
12,12.76,15.0,0.033440002,0.3355847100000006,-1.0,10,0.049600873,-191853.56,-203713.90625
12,12.76,15.0,0.030399993,0.3386247190000006,-1.0,10,0.049600873,25226.638,13001.115234375
12,12.76,15.0,0.027360177,0.3416645350000006,-1.0,10,0.049600873,185829.22,173234.328125
12,12.76,15.0,0.024320183,0.3447045290000006,-1.0,10,0.049600873,279235.44,274720.375
12,12.76,15.0,0.021280171,0.3477445410000006,-1.0,10,0.049600873,329825.92,323887.53125
12,12.76,15.0,0.018240179,0.3507845330000005,-1.0,10,0.049600873,364122.23,358556.6875
12,12.76,15.0,0.015200128,0.3538245840000006,-1.0,10,0.049600873,384300.17,381459.34375
12,12.76,15.0,0.012160197,0.3568645150000006,-1.0,10,0.049600873,400388.32,395957.34375
12,12.76,15.0,0.0091200328,0.3599046792000006,-1.0,10,0.049600873,413906.29,406463.15625
12,12.76,15.0,0.0060800154,0.3629446966000006,-1.0,10,0.049600873,427020.93,420540.9375
12,12.76,15.0,0.0030400341,0.3659846779000005,-1.0,10,0.049600873,441867.15,443863.8125
12,12.76,15.0,-2.9466151e-08,0.3690247414661515,-1.0,10,0.049600873,467305.05,466766.46875
12,12.76,15.0,-0.0030398073,0.3720645193000005,-1.0,10,0.049600873,493069.12,488198.3125
12,12.76,15.0,-0.0060799214,0.3751046334000005,-1.0,10,0.049600873,511127.77,507108.75
12,12.76,15.0,-0.0091199238,0.3781446358000004,-1.0,10,0.049600873,524369.59,522657.34375
12,12.76,15.0,-0.012159879,0.3811845910000004,-1.0,10,0.049600873,532420.82,534003.625
12,12.76,15.0,-0.015199802,0.3842245140000003,-1.0,10,0.049600873,541019.29,542828.5
12,12.76,15.0,-0.018239973,0.3872646850000003,-1.0,10,0.049600873,547981.65,550392.6875
12,12.76,15.0,-0.021279825,0.3903045370000003,-1.0,10,0.049600873,554650.34,554174.8125
12,12.76,15.0,-0.024319844,0.3933445560000003,-1.0,10,0.049600873,561918.26,562579.4375
12,12.76,15.0,-0.027360068,0.3963847800000002,-1.0,10,0.049600873,570107.35,570563.8125
12,12.76,15.0,-0.030399956,0.3994246680000002,-1.0,10,0.049600873,578944.97,577287.5625
12,12.76,15.0,-0.033439981,0.4024646930000003,-1.0,10,0.049600873,592437.88,589054.0625
12,12.76,15.0,-0.036479809,0.4055045210000003,-1.0,10,0.049600873,610454.22,606283.5625
12,12.76,15.0,-0.036480019,0.4079367150000003,1.0,11,0.049600873,463068.5,457941.59375
12,12.76,15.0,-0.03344012,0.4109766140000003,1.0,11,0.049600873,225706.19,223347.015625
12,12.76,15.0,-0.030400175,0.4140165590000003,1.0,11,0.049600873,7043.9335,2541.2734375
12,12.76,15.0,-0.027360097,0.4170566370000004,1.0,11,0.049600873,-157356.01,-161060.34375
12,12.76,15.0,-0.024320163,0.4200965710000004,1.0,11,0.049600873,-263000.19,-268219.5625
12,12.76,15.0,-0.021280125,0.4231366090000004,1.0,11,0.049600873,-320008.19,-327262.1875
12,12.76,15.0,-0.018240039,0.4261766950000004,1.0,11,0.049600873,-360810.72,-367394.34375
12,12.76,15.0,-0.015200034,0.4292167000000004,1.0,11,0.049600873,-386023.9,-391137.46875
12,12.76,15.0,-0.012160102,0.4322566320000004,1.0,11,0.049600873,-403242.73,-408577.09375
12,12.76,15.0,-0.0091200464,0.4352966876000004,1.0,11,0.049600873,-417838.67,-421604.28125
12,12.76,15.0,-0.0060800275,0.4383367065000004,1.0,11,0.049600873,-431402.56,-434211.25
12,12.76,15.0,-0.0030402566,0.4413764774000004,1.0,11,0.049600873,-448015.39,-451020.53125
12,12.76,15.0,-1.9112045e-07,0.4444165428795504,1.0,11,0.049600873,-472109.3,-474133.3125
12,12.76,15.0,0.0030397883,0.4474565223000004,1.0,11,0.049600873,-495599.36,-498927.0
12,12.76,15.0,0.0060799065,0.4504966405000005,1.0,11,0.049600873,-517688.49,-519098.15625
12,12.76,15.0,0.0091197493,0.4535364833000005,1.0,11,0.049600873,-532625.2,-533386.0
12,12.76,15.0,0.012159963,0.4565766970000005,1.0,11,0.049600873,-543066.92,-543051.375
12,12.76,15.0,0.01519978,0.4596165140000005,1.0,11,0.049600873,-550963.29,-550615.5625
12,12.76,15.0,0.018239949,0.4626566830000005,1.0,11,0.049600873,-558348.82,-557339.25
12,12.76,15.0,0.021279804,0.4656965380000005,1.0,11,0.049600873,-565263.41,-563642.75
12,12.76,15.0,0.024319738,0.4687364720000005,1.0,11,0.049600873,-571708.2,-570366.4375
12,12.76,15.0,0.027359846,0.4717765800000005,1.0,11,0.049600873,-577732.36,-578350.875
12,12.76,15.0,0.030399941,0.4748166750000005,1.0,11,0.049600873,-583317.66,-588856.6875
12,12.76,15.0,0.033439825,0.4778565590000005,1.0,11,0.049600873,-593954.67,-605245.6875
12,12.76,15.0,0.036479733,0.4808964670000005,1.0,11,0.049600873,-609407.21,-613650.375
12,12.76,15.0,0.03648025,0.4833287640000005,-1.0,12,0.049600873,-460481.68,-488000.96875
12,12.76,15.0,0.033440032,0.4863689820000004,-1.0,12,0.049600873,-223058.08,-238172.953125
12,12.76,15.0,0.030400173,0.4894088410000004,-1.0,12,0.049600873,-5716.2657,-16658.083984375
12,12.76,15.0,0.027360033,0.4924489810000004,-1.0,12,0.049600873,162378.95,152012.609375
12,12.76,15.0,0.024320095,0.4954889190000004,-1.0,12,0.049600873,266420.04,264634.8125
12,12.76,15.0,0.021280076,0.4985289380000005,-1.0,12,0.049600873,324373.03,321576.25
12,12.76,15.0,0.018240164,0.5015688500000005,-1.0,12,0.049600873,367237.37,362548.90625
12,12.76,15.0,0.015200207,0.5046088070000004,-1.0,12,0.049600873,391467.38,389443.75
12,12.76,15.0,0.012160181,0.5076488330000003,-1.0,12,0.049600873,409011.65,405622.6875
12,12.76,15.0,0.0091200561,0.5106889579000002,-1.0,12,0.049600873,423794.39,416548.71875
12,12.76,15.0,0.0060800342,0.5137289798000003,-1.0,12,0.049600873,437346.37,434198.46875
12,12.76,15.0,0.0030402696,0.5167687444000002,-1.0,12,0.049600873,455000.98,457101.125
12,12.76,15.0,5.9955722e-08,0.5198089540442782,-1.0,12,0.049600873,479590.99,479373.4375
12,12.76,15.0,-0.0030399466,0.5228489606000002,-1.0,12,0.049600873,502940.45,499964.8125
12,12.76,15.0,-0.0060798885,0.5258889025000002,-1.0,12,0.049600873,525003.52,518455.03125
12,12.76,15.0,-0.0091197249,0.5289287389000003,-1.0,12,0.049600873,535540.66,533163.1875
12,12.76,15.0,-0.012159781,0.5319687950000003,-1.0,12,0.049600873,547962.42,544089.1875
12,12.76,15.0,-0.015199751,0.5350087650000003,-1.0,12,0.049600873,556303.83,553334.3125
12,12.76,15.0,-0.018239882,0.5380488960000003,-1.0,12,0.049600873,563567.52,561318.75
12,12.76,15.0,-0.021279883,0.5410888970000003,-1.0,12,0.049600873,570262.0,566361.5
12,12.76,15.0,-0.024319736,0.5441287500000003,-1.0,12,0.049600873,576530.66,573925.6875
12,12.76,15.0,-0.027359955,0.5471689690000002,-1.0,12,0.049600873,582382.23,583170.8125
12,12.76,15.0,-0.030399936,0.5502089500000004,-1.0,12,0.049600873,588198.81,588213.5625
12,12.76,15.0,-0.03343981,0.5532488240000002,-1.0,12,0.049600873,598624.64,601661.0
12,12.76,15.0,-0.036479724,0.5562887380000002,-1.0,12,0.049600873,613595.86,617629.8125
12,12.76,15.0,-0.036480496,0.5587214920000003,1.0,13,0.049600873,464600.85,462564.15625
12,12.76,15.0,-0.033440329,0.5617616590000004,1.0,13,0.049600873,227961.12,225133.0
12,12.76,15.0,-0.030400406,0.5648015820000004,1.0,13,0.049600873,9480.0309,2396.818359375
12,12.76,15.0,-0.027360377,0.5678416110000004,1.0,13,0.049600873,-155286.04,-161585.640625
12,12.76,15.0,-0.024320299,0.5708816890000005,1.0,13,0.049600873,-262235.41,-269690.375
12,12.76,15.0,-0.021280255,0.5739217330000005,1.0,13,0.049600873,-322505.07,-332304.96875
12,12.76,15.0,-0.018240342,0.5769616460000005,1.0,13,0.049600873,-362857.43,-372227.03125
12,12.76,15.0,-0.015200498,0.5800014900000005,1.0,13,0.049600873,-388477.74,-395970.125
12,12.76,15.0,-0.012160269,0.5830417190000005,1.0,13,0.049600873,-405829.36,-413199.65625
12,12.76,15.0,-0.0091204159,0.5860815721000003,1.0,13,0.049600873,-420567.66,-427067.3125
12,12.76,15.0,-0.0060802606,0.5891217274000004,1.0,13,0.049600873,-434135.21,-441775.4375
12,12.76,15.0,-0.0030403114,0.5921616766000004,1.0,13,0.049600873,-450984.84,-461526.34375
12,12.76,15.0,-3.0057512e-07,0.5952016874248803,1.0,13,0.049600873,-475186.6,-487160.5
12,12.76,15.0,0.0039996692,0.5992016572000003,1.0,13,0.049600873,-514717.29,-516576.78125
12,12.76,15.0,0.0079994688,0.6032014568000003,1.0,13,0.049600873,-540775.47,-540529.9375
12,12.76,15.0,0.011999598,0.6072015860000002,1.0,13,0.049600873,-555340.36,-556919.0
12,12.76,15.0,0.015999627,0.6112016150000001,1.0,13,0.049600873,-565367.69,-568685.5
12,12.76,15.0,0.019999723,0.6152017110000002,1.0,13,0.049600873,-574569.66,-578350.875
12,12.76,15.0,0.023999639,0.6192016270000003,1.0,13,0.049600873,-583024.39,-587175.75
12,12.76,15.0,0.027999501,0.6232014890000002,1.0,13,0.049600873,-590731.94,-596420.8125
12,12.76,15.0,0.031999765,0.627201753,1.0,13,0.049600873,-600230.8,-606926.625
12,12.76,15.0,0.035999696,0.631201684,1.0,13,0.049600873,-617494.86,-619953.8125
12,12.76,15.0,0.039999655,0.635201643,1.0,13,0.049600873,-639410.14,-637603.5625
12,12.76,15.0,0.043999739,0.6392017269999999,1.0,13,0.049600873,-661651.31,-665338.9375
12,12.76,15.0,0.047999672,0.6432016599999999,1.0,13,0.049600873,-683089.49,-683829.125
12,12.76,15.0,0.048000263,0.6464022529999998,-1.0,14,0.049600873,-481691.4,-494724.6875
12,12.76,15.0,0.044000273,0.6504022429999998,-1.0,14,0.049600873,-180494.6,-198250.890625
12,12.76,15.0,0.040000117,0.6544023989999999,-1.0,14,0.049600873,75457.377,66495.3515625
12,12.76,15.0,0.036000439,0.6584020769999999,-1.0,14,0.049600873,243877.78,234903.390625
12,12.76,15.0,0.03200013,0.6624023859999999,-1.0,14,0.049600873,337475.58,330611.25
12,12.76,15.0,0.028000204,0.666402312,-1.0,14,0.049600873,386375.27,372214.21875
12,12.76,15.0,0.024000195,0.6704023210000001,-1.0,14,0.049600873,396430.96,401630.46875
12,12.76,15.0,0.020000311,0.6744022050000001,-1.0,14,0.049600873,432094.72,422642.09375
12,12.76,15.0,0.016000357,0.6784021590000001,-1.0,14,0.049600873,445929.2,438610.90625
12,12.76,15.0,0.012000254,0.6824022620000001,-1.0,14,0.049600873,459358.06,451217.875
12,12.76,15.0,0.0080002725,0.6864022435,-1.0,14,0.049600873,473311.71,464245.0625
12,12.76,15.0,0.0040003923,0.6904021237,-1.0,14,0.049600873,495504.89,492820.84375
12,12.76,15.0,2.6998025e-07,0.6944022460197499,-1.0,14,0.049600873,526580.21,519715.71875
12,12.76,15.0,-0.0039997475,0.6984022634999999,-1.0,14,0.049600873,555031.34,544509.4375
12,12.76,15.0,-0.0079996272,0.7024021431999998,-1.0,14,0.049600873,573448.41,565941.3125
12,12.76,15.0,-0.011999714,0.7064022299999998,-1.0,14,0.049600873,584837.34,582750.5625
12,12.76,15.0,-0.015999607,0.7104021229999997,-1.0,14,0.049600873,593448.03,593676.625
12,12.76,15.0,-0.019999617,0.7144021329999997,-1.0,14,0.049600873,601259.65,601661.0
12,12.76,15.0,-0.023999704,0.7184022199999996,-1.0,14,0.049600873,608408.14,609645.4375
12,12.76,15.0,-0.027999744,0.7224022599999996,-1.0,14,0.049600873,614959.12,617209.625
12,12.76,15.0,-0.031999882,0.7264023979999996,-1.0,14,0.049600873,626413.85,628135.625
12,12.76,15.0,-0.035999895,0.7304024109999997,-1.0,14,0.049600873,646468.37,645365.1875
12,12.76,15.0,-0.039999781,0.7344022969999996,-1.0,14,0.049600873,667991.17,662174.4375
12,12.76,15.0,-0.043999775,0.7384022909999997,-1.0,14,0.049600873,688749.21,682765.8125
12,12.76,15.0,-0.047999663,0.7424021789999997,-1.0,14,0.049600873,708202.85,702516.75
12,12.76,15.0,-0.047999834,0.7456023639999997,1.0,15,0.049600873,505830.87,507949.21875
12,12.76,15.0,-0.043999663,0.7496025349999996,1.0,15,0.049600873,207242.51,201284.828125
12,12.76,15.0,-0.03999998,0.7536022179999996,1.0,15,0.049600873,-49058.703,-57420.6015625
12,12.76,15.0,-0.035999884,0.7576023139999997,1.0,15,0.049600873,-222970.07,-232079.578125
12,12.76,15.0,-0.031999825,0.7616023729999997,1.0,15,0.049600873,-325822.82,-328102.65625
12,12.76,15.0,-0.027999827,0.7656023709999996,1.0,15,0.049600873,-382731.37,-385044.09375
12,12.76,15.0,-0.023999658,0.7696025399999996,1.0,15,0.049600873,-406486.97,-416561.5
12,12.76,15.0,-0.019999727,0.7736024709999997,1.0,15,0.049600873,-432177.32,-437152.875
12,12.76,15.0,-0.015999797,0.7776024009999997,1.0,15,0.049600873,-446639.43,-453541.9375
12,12.76,15.0,-0.011999825,0.7816023729999997,1.0,15,0.049600873,-461045.37,-466569.125
12,12.76,15.0,-0.0079997508,0.7856024471999997,1.0,15,0.049600873,-474888.52,-479596.34375
12,12.76,15.0,-0.0039997608,0.7896024371999998,1.0,15,0.049600873,-493664.15,-497246.09375
12,12.76,15.0,5.4958764e-08,0.7936022529587639,1.0,15,0.049600873,-523867.13,-524561.125
12,12.76,15.0,0.0040001561,0.7976023540999999,1.0,15,0.049600873,-553568.27,-555238.0625
12,12.76,15.0,0.0080000114,0.8016022093999998,1.0,15,0.049600873,-577551.79,-577930.625
12,12.76,15.0,0.012000139,0.8056023369999998,1.0,15,0.049600873,-589527.88,-591798.3125
12,12.76,15.0,0.01600031,0.8096025079999999,1.0,15,0.049600873,-598333.89,-601463.625
12,12.76,15.0,0.020000335,0.813602533,1.0,15,0.049600873,-606542.93,-609027.8125
12,12.76,15.0,0.024000024,0.8176022219999999,1.0,15,0.049600873,-614009.11,-616592.0
12,12.76,15.0,0.028000216,0.8216024139999999,1.0,15,0.049600873,-620803.52,-625416.875
12,12.76,15.0,0.032000283,0.8256024809999999,1.0,15,0.049600873,-630142.1,-636763.125
12,12.76,15.0,0.036000054,0.8296022519999999,1.0,15,0.049600873,-647997.25,-651471.25
12,12.76,15.0,0.040000145,0.833602343,1.0,15,0.049600873,-668313.7,-672482.875
12,12.76,15.0,0.044000247,0.837602445,1.0,15,0.049600873,-686645.37,-701478.875
12,12.76,15.0,0.048000322,0.84160252,1.0,15,0.049600873,-703534.22,-723330.9375
12,12.76,15.0,0.048000854,0.8448030900000001,-1.0,16,0.049600873,-500365.79,-547673.9375
12,12.76,15.0,0.04400058,0.8488033640000001,-1.0,16,0.049600873,-201116.23,-234180.734375
12,12.76,15.0,0.040000821,0.8528031230000002,-1.0,16,0.049600873,55310.017,33848.5703125
12,12.76,15.0,0.03600071,0.8568032340000001,-1.0,16,0.049600873,229331.02,214627.1875
12,12.76,15.0,0.032000765,0.8608031790000001,-1.0,16,0.049600873,331677.59,328089.84375
12,12.76,15.0,0.02800054,0.8648034040000001,-1.0,16,0.049600873,385866.75,374105.28125
12,12.76,15.0,0.024000578,0.8688033660000002,-1.0,16,0.049600873,396132.43,405832.8125
12,12.76,15.0,0.020000836,0.8728031080000002,-1.0,16,0.049600873,434149.53,428105.09375
12,12.76,15.0,0.016000847,0.8768030970000003,-1.0,16,0.049600873,450973.87,444494.15625
12,12.76,15.0,0.012000638,0.8808033060000001,-1.0,16,0.049600873,464209.68,457521.34375
12,12.76,15.0,0.0080008446,0.8848030994,-1.0,16,0.049600873,478635.93,472229.46875
12,12.76,15.0,0.0040005462,0.8888033978000001,-1.0,16,0.049600873,500146.16,502486.1875
12,12.76,15.0,5.7932109e-07,0.8928033646789101,-1.0,16,0.049600873,531866.08,530221.5625
12,12.76,15.0,-0.0039991344,0.8968030784000001,-1.0,16,0.049600873,561139.34,555855.6875
12,12.76,15.0,-0.007999294,0.9008032380000002,-1.0,16,0.049600873,596325.5,577287.5625
12,12.76,15.0,-0.011999188,0.9048031320000002,-1.0,16,0.049600873,608322.15,592836.125
12,12.76,15.0,-0.015999423,0.9088033670000002,-1.0,16,0.049600873,619836.34,602501.5
12,12.76,15.0,-0.01999925,0.9128031940000002,-1.0,16,0.049600873,628778.21,611326.375
12,12.76,15.0,-0.02399934,0.9168032840000002,-1.0,16,0.049600873,636547.35,622672.625
12,12.76,15.0,-0.027999209,0.9208031530000002,-1.0,16,0.049600873,653692.96,635279.5625
12,12.76,15.0,-0.031999329,0.924803273,-1.0,16,0.049600873,680645.01,652088.875
12,12.76,15.0,-0.035999384,0.928803328,-1.0,16,0.049600873,707095.92,675201.625
12,12.76,15.0,-0.039999118,0.9328030620000002,-1.0,16,0.049600873,731629.97,697894.1875
12,12.76,15.0,-0.043999166,0.9368031100000002,-1.0,16,0.049600873,754568.35,723948.5625
12,12.76,15.0,-0.047999125,0.940803069,-1.0,16,0.049600873,777012.69,743699.5
12,12.76,15.0,-0.047999564,0.944003528,1.0,17,0.049600873,578641.48,536945.25
12,12.76,15.0,-0.043999591,0.9480035010000002,1.0,17,0.049600873,265989.46,218934.578125
12,12.76,15.0,-0.039999705,0.9520033870000002,1.0,17,0.049600873,83.169636,-37774.75
12,12.76,15.0,-0.035999569,0.9560035230000002,1.0,17,0.049600873,-183911.21,-217791.6875
12,12.76,15.0,-0.031999679,0.9600034130000002,1.0,17,0.049600873,-301092.13,-326421.71875
12,12.76,15.0,-0.02799965,0.9640034420000004,1.0,17,0.049600873,-379081.18,-388826.1875
12,12.76,15.0,-0.023999482,0.9680036100000003,1.0,17,0.049600873,-411965.55,-419923.375
12,12.76,15.0,-0.019999661,0.9720034310000004,1.0,17,0.049600873,-434395.61,-440514.75
12,12.76,15.0,-0.01599982,0.9760032720000006,1.0,17,0.049600873,-449550.76,-456483.5625
12,12.76,15.0,-0.011999574,0.9800035180000006,1.0,17,0.049600873,-464618.85,-469510.75
12,12.76,15.0,-0.0079994856,0.9840036064000006,1.0,17,0.049600873,-482208.3,-483798.65625
12,12.76,15.0,-0.0039997777,0.9880033143000008,1.0,17,0.049600873,-511776.54,-504390.03125
12,12.76,15.0,3.5415353e-07,0.9920034461535306,1.0,17,0.049600873,-542005.43,-535907.4375
Code/reports/hysteretic_curves/3W/H45_B29/case_22_with_rnn_preds.csv
This source diff could not be displayed because it is too large. You can view the blob instead.
Code/reports/hysteretic_curves/3W/H45_B34/case_21_with_rnn_preds.csv
This source diff could not be displayed because it is too large. You can view the blob instead.
Code/reports/hysteretic_curves/5W/H60_B34/case_70_with_rnn_preds.csv
This source diff could not be displayed because it is too large. You can view the blob instead.
Code/src/hysteretic_curves/hysteretic_curves.py
import os import os
import pandas as pd
import numpy as np import numpy as np
import pandas as pd
# import matplotlib.pyplot as plt # import matplotlib.pyplot as plt
# Configuration constants and bounds # Configuration constants and bounds
W = 5 # Number of windows W = 2 # Number of windows
B = 34 # Width identifier B = 29 # Width identifier
H = 30 H = 30
if W == 2: if W == 2:
H = 30 # Adjust height identifier for two window case H = 30 # Adjust height identifier for two window case
...@@ -40,22 +42,56 @@ print(f"Number of cases: {N_CASES}") ...@@ -40,22 +42,56 @@ print(f"Number of cases: {N_CASES}")
print(f"Number of windows per case: {n_windows}") print(f"Number of windows per case: {n_windows}")
# Column names for tw1, tw2, ... # Column names for tw1, tw2, ...
thickness_cols = [f"tw{i+1}" for i in range(n_windows)] thickness_cols = [f"tw{i + 1}" for i in range(n_windows)]
all_rows = [] all_rows = []
def load_tfdmap(case_folder_path):
tfdmap_file = os.path.join(case_folder_path, "TFDMap.txt")
if not os.path.isfile(tfdmap_file):
raise FileNotFoundError(f"TFDMap file not found: {tfdmap_file}")
df_tfdmap = pd.read_csv(
tfdmap_file,
comment="#",
sep=r"\s+",
engine="python",
header=None,
)
# Some files contain an extra non-comment header row such as
# "Tiempo promW1 ... promF0". Keep only fully numeric rows.
df_tfdmap = df_tfdmap.apply(pd.to_numeric, errors="coerce")
df_tfdmap = df_tfdmap.dropna(axis=0, how="any").reset_index(drop=True)
if df_tfdmap.empty:
raise ValueError(f"TFDMap file has no numeric data rows: {tfdmap_file}")
if df_tfdmap.shape[1] < 2:
raise ValueError(
f"TFDMap file must contain at least two columns after ignoring headers: {tfdmap_file}"
)
df_tfdmap = df_tfdmap.iloc[:, 1:].copy()
n_cols = df_tfdmap.shape[1]
if n_cols == 1:
df_tfdmap.columns = ["TFDMapF"]
else:
df_tfdmap.columns = [f"TFDMapW{i + 1}" for i in range(n_cols - 1)] + ["TFDMapF"]
return df_tfdmap
# Detect all numeric folders automatically # Detect all numeric folders automatically
case_folders = sorted( case_folders = sorted([f for f in os.listdir(CURVES_DIR) if f.isdigit()], key=int)
[f for f in os.listdir(CURVES_DIR) if f.isdigit()],
key=int
)
case_folders_sorted = sorted(case_folders, key=lambda x: int(x)) case_folders_sorted = sorted(case_folders, key=lambda x: int(x))
# thickness_data debe tener la MISMA longitud y orden que case_folders_sorted # thickness_data debe tener la MISMA longitud y orden que case_folders_sorted
thickness_by_case = { thickness_by_case = {
int(folder): thickness_data[idx] int(folder): thickness_data[idx] for idx, folder in enumerate(case_folders_sorted)
for idx, folder in enumerate(case_folders_sorted)
} }
# === Loop === # === Loop ===
...@@ -67,7 +103,13 @@ for folder_name in case_folders_sorted: ...@@ -67,7 +103,13 @@ for folder_name in case_folders_sorted:
print(f"⚠️ Folder not found: {folder_path}, skipping.") print(f"⚠️ Folder not found: {folder_path}, skipping.")
continue continue
files = [f for f in os.listdir(folder_path) if f.endswith((".txt", ".dat", ".csv"))] files = [
f
for f in os.listdir(folder_path)
if f.endswith((".txt", ".dat", ".csv"))
and f != "TFDMap.txt"
and not f.startswith("merged_dataset_")
]
if not files: if not files:
print(f"⚠️ No hysteresis file in {folder_path}.") print(f"⚠️ No hysteresis file in {folder_path}.")
continue continue
...@@ -76,8 +118,12 @@ for folder_name in case_folders_sorted: ...@@ -76,8 +118,12 @@ for folder_name in case_folders_sorted:
print(f"Processing Case {i}: {datafile}") print(f"Processing Case {i}: {datafile}")
df_curve = pd.read_csv( df_curve = pd.read_csv(
datafile, comment="#", sep=r"\s+", engine="python", datafile,
names=["Displ", "Force"], header=None comment="#",
sep=r"\s+",
engine="python",
names=["Displ", "Force"],
header=None,
) )
df_curve["Case"] = i df_curve["Case"] = i
...@@ -116,9 +162,8 @@ def compute_load_dir(displ): ...@@ -116,9 +162,8 @@ def compute_load_dir(displ):
df["LoadDir"] = df.groupby("Case")["Displ"].transform(compute_load_dir) df["LoadDir"] = df.groupby("Case")["Displ"].transform(compute_load_dir)
# CumDispl: cumulative |ΔΔ| per case # CumDispl: cumulative |ΔΔ| per case
df["CumDispl"] = ( df["CumDispl"] = df.groupby("Case")["Displ"].transform(
df.groupby("Case")["Displ"] lambda s: s.diff().abs().fillna(0).cumsum()
.transform(lambda s: s.diff().abs().fillna(0).cumsum())
) )
...@@ -143,18 +188,47 @@ df["MaxAmpl"] = df.groupby("Case")["Displ"].transform(lambda x: x.abs().max()) ...@@ -143,18 +188,47 @@ df["MaxAmpl"] = df.groupby("Case")["Displ"].transform(lambda x: x.abs().max())
# 4) Reorder point-level columns (Force last) # 4) Reorder point-level columns (Force last)
# ------------------------------------------------------------------ # ------------------------------------------------------------------
ordered_cols_point = ( ordered_cols_point = (
["Case"] + ["Case"]
thickness_cols + + thickness_cols
["Displ", "CumDispl", "LoadDir", "CycleNum", "MaxAmpl", "Force"] + ["Displ", "CumDispl", "LoadDir", "CycleNum", "MaxAmpl", "Force"]
) )
df = df[ordered_cols_point] df = df[ordered_cols_point]
print("\n=== Point-level DataFrame (df) ===") # Point-level export: keep 1 out of every 5 rows within each case
print(df.head()) keep_mask = df.groupby("Case", sort=False).cumcount() % 5 == 0
df_points = df.loc[keep_mask].reset_index(drop=True)
# Add TFDMap data from the corresponding file in each case folder
tfdmap_frames = []
for folder_name in case_folders_sorted:
case_id = int(folder_name)
folder_path = os.path.join(CURVES_DIR, folder_name)
df_case_points = df_points[df_points["Case"] == case_id].reset_index(drop=True)
if df_case_points.empty:
continue
df_tfdmap = load_tfdmap(folder_path)
if len(df_tfdmap) != len(df_case_points):
raise ValueError(
f"TFDMap length mismatch for case {case_id}: "
f"{len(df_tfdmap)} rows in TFDMap.txt vs {len(df_case_points)} sampled rows."
)
df_case_points = pd.concat(
[df_case_points, df_tfdmap.reset_index(drop=True)],
axis=1,
)
tfdmap_frames.append(df_case_points)
df_points = pd.concat(tfdmap_frames, ignore_index=True)
print("\n=== Point-level DataFrame (df_points) ===")
# Save point-level dataset # Save point-level dataset
output_file_points = os.path.join(CURVES_DIR, "merged_dataset_points.csv") output_file_points = os.path.join(CURVES_DIR, "merged_dataset_points.csv")
df.to_csv(output_file_points, index=False) df_points.to_csv(output_file_points, index=False)
print(f"\nSaved point-level dataset to: {output_file_points}") print(f"\nSaved point-level dataset to: {output_file_points}")
# ================================================================== # ==================================================================
...@@ -238,21 +312,25 @@ df_cycles["StiffnessRatio"] = df_cycles.groupby("Case")["SecantStiffness"].trans ...@@ -238,21 +312,25 @@ df_cycles["StiffnessRatio"] = df_cycles.groupby("Case")["SecantStiffness"].trans
# Order cycle-level columns nicely # Order cycle-level columns nicely
ordered_cols_cycles = ( ordered_cols_cycles = (
["Case", "CycleNum"] + ["Case", "CycleNum"]
thickness_cols + + thickness_cols
[ + [
"CycleMaxAmpl", "CycleMaxAmpl",
"PosEnvDispl", "PosEnvForce", "PosEnvDispl",
"NegEnvDispl", "NegEnvForce", "PosEnvForce",
"LoopEnergySigned", "LoopEnergyAbs", "CumLoopEnergyAbs", "NegEnvDispl",
"SecantStiffness", "StiffnessRatio" "NegEnvForce",
"LoopEnergySigned",
"LoopEnergyAbs",
"CumLoopEnergyAbs",
"SecantStiffness",
"StiffnessRatio",
] ]
) )
df_cycles = df_cycles[ordered_cols_cycles] df_cycles = df_cycles[ordered_cols_cycles]
print("\n=== Cycle-level DataFrame (df_cycles) ===") print("\n=== Cycle-level DataFrame (df_cycles) ===")
print(df_cycles.head())
# Save cycle-level summary # Save cycle-level summary
output_file_cycles = os.path.join(CURVES_DIR, "merged_dataset_cycles.csv") output_file_cycles = os.path.join(CURVES_DIR, "merged_dataset_cycles.csv")
......
Code/src/hysteretic_curves/predict_hysteretic_curves.py
...@@ -127,7 +127,7 @@ def make_windows_for_case_fast( ...@@ -127,7 +127,7 @@ def make_windows_for_case_fast(
CaseWindows = Dict[ CaseWindows = Dict[
int, Tuple[np.ndarray, np.ndarray, np.ndarray] int, Tuple[np.ndarray, np.ndarray, np.ndarray]
] # case -> (x, y, idx) ] # case -> (x, y, idx)
WINDOW_CACHE: Dict[int, CaseWindows] = {} # window_size -> cache WINDOW_CACHE: Dict[Tuple[str, int], CaseWindows] = {} # (target_col, window_size) -> cache
def precompute_windows_for_cases( def precompute_windows_for_cases(
...@@ -154,15 +154,16 @@ def get_cache_for_window_size( ...@@ -154,15 +154,16 @@ def get_cache_for_window_size(
target_col: str, target_col: str,
window_size: int, window_size: int,
) -> CaseWindows: ) -> CaseWindows:
if window_size not in WINDOW_CACHE: cache_key = (target_col, window_size)
WINDOW_CACHE[window_size] = precompute_windows_for_cases( if cache_key not in WINDOW_CACHE:
WINDOW_CACHE[cache_key] = precompute_windows_for_cases(
df=df, df=df,
cases=cases_to_cache, cases=cases_to_cache,
feature_cols=feature_cols, feature_cols=feature_cols,
target_col=target_col, target_col=target_col,
window_size=window_size, window_size=window_size,
) )
return WINDOW_CACHE[window_size] return WINDOW_CACHE[cache_key]
def build_from_cache( def build_from_cache(
...@@ -493,6 +494,23 @@ def strip_compile_prefix(state_dict: dict) -> dict: ...@@ -493,6 +494,23 @@ def strip_compile_prefix(state_dict: dict) -> dict:
return {k.replace("_orig_mod.", "", 1): v for k, v in state_dict.items()} return {k.replace("_orig_mod.", "", 1): v for k, v in state_dict.items()}
def target_sort_key(col_name: str) -> Tuple[int, int]:
if col_name == "Force":
return (0, 0)
if col_name.startswith("TFDMapW"):
try:
return (1, int(col_name.replace("TFDMapW", "")))
except ValueError:
return (1, 9999)
if col_name == "TFDMapF":
return (2, 0)
return (3, 0)
def target_prediction_col(target_col: str) -> str:
return f"{target_col}_RNN"
# ------------------------- # -------------------------
# Final training on 100% train_pool (cached) - no val # Final training on 100% train_pool (cached) - no val
# Keep best by TRAIN loss # Keep best by TRAIN loss
...@@ -688,7 +706,6 @@ def main(): ...@@ -688,7 +706,6 @@ def main():
model_dir = f"../../models/hysteretic_curves/{w_val}W/models" model_dir = f"../../models/hysteretic_curves/{w_val}W/models"
os.makedirs(model_dir, exist_ok=True) os.makedirs(model_dir, exist_ok=True)
model_path = os.path.join(model_dir, f"B{b_val}_H{h_val}.joblib")
n_train_case = 0 n_train_case = 0
predict_case = 0 predict_case = 0
...@@ -727,19 +744,17 @@ def main(): ...@@ -727,19 +744,17 @@ def main():
df = pd.read_csv(data_file) df = pd.read_csv(data_file)
# Keep only 1 out of every 5 rows within each case.
keep_mask = df.groupby("Case", sort=False).cumcount() % 5 == 0
df = df.loc[keep_mask].reset_index(drop=True)
thickness_cols = [c for c in df.columns if c.startswith("tw")] thickness_cols = [c for c in df.columns if c.startswith("tw")]
feature_cols = [ feature_cols = [
"Displ", "Displ",
"CumDispl", "CumDispl",
# "LoadDir",
# "CycleNum",
# "MaxAmpl",
] + thickness_cols ] + thickness_cols
target_col = "Force" target_cols = sorted(
[c for c in df.columns if c == "Force" or c.startswith("TFDMap")],
key=target_sort_key,
)
if not target_cols:
raise RuntimeError("No target columns found. Expected Force and/or TFDMap* columns.")
all_cases = sorted(df["Case"].unique()) all_cases = sorted(df["Case"].unique())
train_pool = [int(c) for c in all_cases if c <= n_train_case] train_pool = [int(c) for c in all_cases if c <= n_train_case]
...@@ -757,6 +772,8 @@ def main(): ...@@ -757,6 +772,8 @@ def main():
print("Predict (test) case:", predict_case) print("Predict (test) case:", predict_case)
print("Train pool cases (for CV):", train_pool) print("Train pool cases (for CV):", train_pool)
print("n_cases total:", len(all_cases), "| train_pool:", len(train_pool)) print("n_cases total:", len(all_cases), "| train_pool:", len(train_pool))
print("Feature columns:", feature_cols)
print("Target columns:", target_cols)
print("torch.compile CV:", "ON" if (device == "cuda" and compile_in_cv) else "OFF") print("torch.compile CV:", "ON" if (device == "cuda" and compile_in_cv) else "OFF")
print( print(
"torch.compile FINAL:", "torch.compile FINAL:",
...@@ -770,269 +787,266 @@ def main(): ...@@ -770,269 +787,266 @@ def main():
# Cache cases needed (train_pool for CV, plus test case for prediction) # Cache cases needed (train_pool for CV, plus test case for prediction)
cases_to_cache_all = sorted(set(train_pool + [int(predict_case)])) cases_to_cache_all = sorted(set(train_pool + [int(predict_case)]))
# ---- CV Tuning ----
if optuna is None: if optuna is None:
raise ImportError( raise ImportError(
"Optuna is required for Bayesian CV search. Install it with `pip install optuna`." "Optuna is required for Bayesian CV search. Install it with `pip install optuna`."
) )
print(f"Optuna search: TPE sampler | n_trials={n_trials}")
optuna.logging.set_verbosity(optuna.logging.WARNING) optuna.logging.set_verbosity(optuna.logging.WARNING)
sampler = optuna.samplers.TPESampler(seed=SEED) df_test = df[df["Case"] == predict_case].copy()
pruner = optuna.pruners.MedianPruner( df_test_pred = df_test.copy()
n_startup_trials=min(5, n_trials), n_warmup_steps=2 rmse_by_target = {}
)
study = optuna.create_study(direction="minimize", sampler=sampler, pruner=pruner) for target_col in target_cols:
print("\n============================================================")
print(f"TARGET: {target_col}")
print(f"Optuna search: TPE sampler | n_trials={n_trials}")
sampler = optuna.samplers.TPESampler(seed=SEED)
pruner = optuna.pruners.MedianPruner(
n_startup_trials=min(5, n_trials), n_warmup_steps=2
)
study = optuna.create_study(
direction="minimize", sampler=sampler, pruner=pruner
)
def objective(trial) -> float:
params = suggest_params(trial, device=device)
ws = int(params["window_size"])
params_model = dict(params)
params_model.pop("window_size", None)
cache_ws = get_cache_for_window_size(
df=df,
cases_to_cache=cases_to_cache_all,
feature_cols=feature_cols,
target_col=target_col,
window_size=ws,
)
feature_dim = None
for c in cases_to_cache_all:
x_c = cache_ws[int(c)][0]
if x_c.shape[0] > 0:
feature_dim = x_c.shape[-1]
break
if feature_dim is None:
raise RuntimeError(
"All cases have 0 windows for this window_size. Lower window_size."
)
mean_rmse, std_rmse, fold_rmses = cv_score_for_params_cached(
cache_ws=cache_ws,
feature_dim=feature_dim,
folds=folds,
params_model=params_model,
max_epochs=max_epachs_tune,
patience=patience_tune,
grad_clip=1.0,
device=device,
compile_model=(compile_in_cv and device == "cuda"),
trial=trial,
)
if fold_rmses is None:
print(f"[Trial {trial.number + 1:02d}] OOM -> pruning trial | {params}")
raise optuna.TrialPruned()
fold_str = ", ".join([f"{v:.3f}" for v in fold_rmses.tolist()])
print(
f"[{target_col}][Trial {trial.number + 1:02d}]"
f" CV mean RMSE={mean_rmse:.4f} | std={std_rmse:.4f} | folds=[{fold_str}]"
)
print(f" params: {params}")
def objective(trial) -> float: trial.set_user_attr("std_rmse", float(std_rmse))
params = suggest_params(trial, device=device) trial.set_user_attr("params_full", params)
# Separate window_size from model params (cached funcs must NOT receive window_size) gc.collect()
ws = int(params["window_size"]) if torch.cuda.is_available():
params_model = dict(params) torch.cuda.empty_cache()
params_model.pop("window_size", None)
return float(mean_rmse)
study.optimize(objective, n_trials=n_trials, show_progress_bar=False)
completed_trials = [
trial
for trial in study.trials
if trial.state == optuna.trial.TrialState.COMPLETE
]
if not completed_trials:
raise RuntimeError(
f"No valid Optuna trial found for target {target_col}. "
"Try lowering window sizes or check your data."
)
# Window cache for this ws (build once, reuse across folds) best_trial = min(
cache_ws = get_cache_for_window_size( completed_trials,
key=lambda trial: (
float(trial.value),
float(trial.user_attrs.get("std_rmse", np.inf)),
),
)
best = {
"mean": float(best_trial.value),
"std": float(best_trial.user_attrs.get("std_rmse", np.inf)),
"params": best_trial.user_attrs.get("params_full"),
}
print("\n====================")
print(f"BEST (OPTUNA CV TUNING) - {target_col}")
print("CV mean RMSE:", best["mean"])
print("CV std RMSE :", best["std"])
print("params:", best["params"])
print("====================\n")
if best["params"] is None:
raise RuntimeError(
f"No valid trial found for target {target_col}. "
"Try lowering window sizes or check your data."
)
best_params = cast(dict, best["params"])
ws_best = int(best_params["window_size"])
best_params_model = dict(best_params)
best_params_model.pop("window_size", None)
cache_ws_best = get_cache_for_window_size(
df=df, df=df,
cases_to_cache=cases_to_cache_all, cases_to_cache=cases_to_cache_all,
feature_cols=feature_cols, feature_cols=feature_cols,
target_col=target_col, target_col=target_col,
window_size=ws, window_size=ws_best,
) )
feature_dim = None feature_dim = None
for c in cases_to_cache_all: for c in cases_to_cache_all:
x_c = cache_ws[int(c)][0] x_c = cache_ws_best[int(c)][0]
if x_c.shape[0] > 0: if x_c.shape[0] > 0:
feature_dim = x_c.shape[-1] feature_dim = x_c.shape[-1]
break break
if feature_dim is None: if feature_dim is None:
raise RuntimeError( raise RuntimeError(
"All cases have 0 windows for this window_size. Lower window_size." f"All cases have 0 windows for best window_size in target {target_col}."
) )
mean_rmse, std_rmse, fold_rmses = cv_score_for_params_cached( print("============================================================")
cache_ws=cache_ws, print(f"FINAL TRAIN (100% train_pool, no val) - {target_col}")
print("Train cases:", train_pool)
print("Using best parameters:", best_params)
print("============================================================")
final_state, final_scalers = train_final_full_trainpool_cached(
cache_ws=cache_ws_best,
feature_dim=feature_dim, feature_dim=feature_dim,
folds=folds, train_cases=train_pool,
params_model=params_model, max_epochs=max_epachs_final,
max_epochs=max_epachs_tune,
patience=patience_tune,
grad_clip=1.0, grad_clip=1.0,
device=device, device=device,
compile_model=(compile_in_cv and device == "cuda"), compile_model=(compile_in_final and device == "cuda"),
trial=trial, **best_params_model,
) )
if fold_rmses is None: scaler_x, scaler_y = final_scalers
print(f"[Trial {trial.number + 1:02d}] OOM -> pruning trial | {params}")
raise optuna.TrialPruned() model_path = os.path.join(model_dir, f"B{b_val}_H{h_val}_{target_col}.joblib")
bundle = {
fold_str = ", ".join([f"{v:.3f}" for v in fold_rmses.tolist()]) "framework": "pytorch",
print( "model_class": "LSTMRegressor",
f"[Trial {trial.number + 1:02d}]" "state_dict": final_state,
f" CV mean RMSE={mean_rmse:.4f} | std={std_rmse:.4f} | folds=[{fold_str}]" "params": {**best_params_model, "window_size": ws_best},
) "feature_cols": feature_cols,
print(f" params: {params}") "target_col": target_col,
"scaler_x": scaler_x,
trial.set_user_attr("std_rmse", float(std_rmse)) "scaler_y": scaler_y,
trial.set_user_attr("params_full", params) "meta": {
"W": w_val,
gc.collect() "B": b_val,
if torch.cuda.is_available(): "H": h_val,
torch.cuda.empty_cache() "n_train_case": n_train_case,
"predict_case": predict_case,
return float(mean_rmse) "seed": SEED,
"device_trained": device,
study.optimize(objective, n_trials=n_trials, show_progress_bar=False) },
}
completed_trials = [ joblib.dump(bundle, model_path, compress=3)
trial print(f"Saved best model bundle to: {model_path}")
for trial in study.trials
if trial.state == optuna.trial.TrialState.COMPLETE x_test, y_test, test_indices = cache_ws_best[int(predict_case)]
] if x_test.shape[0] == 0:
if not completed_trials: raise RuntimeError(
raise RuntimeError( f"Case {predict_case} has fewer points ({len(df_test)}) than "
"No valid Optuna trial found. Try lowering window sizes or check your data." f"window_size ({ws_best}) for target {target_col}."
) )
best_trial = min(
completed_trials,
key=lambda trial: (
float(trial.value),
float(trial.user_attrs.get("std_rmse", np.inf)),
),
)
best = {
"mean": float(best_trial.value),
"std": float(best_trial.user_attrs.get("std_rmse", np.inf)),
"params": best_trial.user_attrs.get("params_full"),
}
print("\n====================")
print("BEST (OPTUNA CV TUNING)")
print("CV mean RMSE:", best["mean"])
print("CV std RMSE :", best["std"])
print("params:", best["params"])
print("====================\n")
if best["params"] is None: x_test_scaled = (
raise RuntimeError( scaler_x.transform(x_test.reshape(-1, feature_dim))
"No valid trial found. Try lowering window sizes or check your data." .reshape(x_test.shape)
.astype(np.float32, copy=False)
) )
best_params = cast(dict, best["params"]) final_model = LSTMRegressor(
ws_best = int(best_params["window_size"]) input_dim=feature_dim,
best_params_model = dict(best_params) hidden_dim=best_params_model["hidden_dim"],
best_params_model.pop("window_size", None) dense_dim=best_params_model["dense_dim"],
num_layers=best_params_model["num_layers"],
# Ensure cache exists for best ws dropout=best_params_model["dropout"],
cache_ws_best = get_cache_for_window_size( ).to(device)
df=df, final_model.load_state_dict(strip_compile_prefix(final_state))
cases_to_cache=cases_to_cache_all, final_model.eval()
feature_cols=feature_cols,
target_col=target_col, test_ds = WindowDataset(
window_size=ws_best, x_test_scaled, np.zeros((x_test_scaled.shape[0],), dtype=np.float32)
)
# feature_dim again
feature_dim = None
for c in cases_to_cache_all:
x_c = cache_ws_best[int(c)][0]
if x_c.shape[0] > 0:
feature_dim = x_c.shape[-1]
break
if feature_dim is None:
raise RuntimeError(
"All cases have 0 windows for best window_size. Lower window_size."
) )
test_loader = make_loader(
# ---- Final train on 100% train_pool (NO VAL) ---- test_ds,
print("============================================================") batch_size=best_params_model["batch_size"],
print("FINAL TRAIN (100% train_pool, no val)") shuffle=False,
print("Train cases:", train_pool) device=device,
print("Using best parameters:", best_params) use_workers=False,
print("============================================================")
final_state, final_scalers = train_final_full_trainpool_cached(
cache_ws=cache_ws_best,
feature_dim=feature_dim,
train_cases=train_pool,
max_epochs=max_epachs_final,
grad_clip=1.0,
device=device,
compile_model=(compile_in_final and device == "cuda"),
**best_params_model,
)
scaler_x, scaler_y = final_scalers
# -------------------------
# Save best model bundle (joblib)
# -------------------------
bundle = {
"framework": "pytorch",
"model_class": "LSTMRegressor",
"state_dict": final_state, # already CPU tensors
"params": {**best_params_model, "window_size": ws_best},
"feature_cols": feature_cols,
"target_col": target_col,
"scaler_x": scaler_x,
"scaler_y": scaler_y,
"meta": {
"W": w_val,
"B": b_val,
"H": h_val,
"n_train_case": n_train_case,
"predict_case": predict_case,
"seed": SEED,
"device_trained": device,
},
}
joblib.dump(bundle, model_path, compress=3)
print(f"Saved best model bundle to: {model_path}")
# ---- Predict predict_case ----
df_test = df[df["Case"] == predict_case].copy()
x_test, y_test, test_indices = cache_ws_best[int(predict_case)]
if x_test.shape[0] == 0:
raise RuntimeError(
f"Case {predict_case} has fewer points ({len(df_test)}) than window_size ({ws_best})."
) )
x_test_scaled = ( use_amp = device == "cuda"
scaler_x.transform(x_test.reshape(-1, feature_dim)) preds = []
.reshape(x_test.shape) with torch.no_grad():
.astype(np.float32, copy=False) for xb, _ in test_loader:
) xb = xb.to(device, non_blocking=True)
if use_amp:
final_model = LSTMRegressor( try:
input_dim=feature_dim, with torch.amp.autocast("cuda", enabled=True):
hidden_dim=best_params_model["hidden_dim"], pb = final_model(xb)
dense_dim=best_params_model["dense_dim"], except Exception:
num_layers=best_params_model["num_layers"], with torch.cuda.amp.autocast(enabled=True):
dropout=best_params_model["dropout"], pb = final_model(xb)
).to(device) else:
pb = final_model(xb)
final_state = strip_compile_prefix(final_state) preds.append(pb.detach().float().cpu().numpy())
# compile not necessary for predict, but harmless if final was compiled; keep OFF to avoid y_pred_scaled = np.concatenate(preds, axis=0).ravel()
# overhead y_pred = scaler_y.inverse_transform(y_pred_scaled.reshape(-1, 1)).ravel()
final_model.load_state_dict(final_state)
final_model.eval()
test_ds = WindowDataset( test_rmse = rmse(y_test, y_pred)
x_test_scaled, np.zeros((x_test_scaled.shape[0],), dtype=np.float32) rmse_by_target[target_col] = test_rmse
) print(f"Test RMSE on Case {predict_case} for {target_col} = {test_rmse:.4f}")
test_loader = make_loader(
test_ds,
batch_size=best_params_model["batch_size"],
shuffle=False,
device=device,
use_workers=False,
)
use_amp = device == "cuda" pred_col = target_prediction_col(target_col)
preds = [] df_test_pred[pred_col] = np.nan
with torch.no_grad(): for idx, pred in zip(test_indices, y_pred):
for xb, _ in test_loader: df_test_pred.loc[idx, pred_col] = pred
xb = xb.to(device, non_blocking=True)
if use_amp:
try:
with torch.amp.autocast("cuda", enabled=True):
pb = final_model(xb)
except Exception:
with torch.cuda.amp.autocast(enabled=True):
pb = final_model(xb)
else:
pb = final_model(xb)
preds.append(pb.detach().float().cpu().numpy())
y_pred_scaled = np.concatenate(preds, axis=0).ravel()
y_pred = scaler_y.inverse_transform(y_pred_scaled.reshape(-1, 1)).ravel()
test_rmse = rmse(y_test, y_pred) del final_model, test_loader, test_ds
print(f"Test RMSE on Case {predict_case} = {test_rmse:.4f}") gc.collect()
if torch.cuda.is_available():
# ---- Save results ---- torch.cuda.empty_cache()
df_test_pred = df_test.copy()
df_test_pred["Force_RNN"] = np.nan
for idx, pred in zip(test_indices, y_pred):
df_test_pred.loc[idx, "Force_RNN"] = pred
out_folder = f"../../reports/hysteretic_curves/{w_val}W/H{h_val}_B{b_val}" out_folder = f"../../reports/hysteretic_curves/{w_val}W/H{h_val}_B{b_val}"
os.makedirs(out_folder, exist_ok=True) os.makedirs(out_folder, exist_ok=True)
out_file = os.path.join(out_folder, f"case_{predict_case}_with_rnn_preds.csv") out_file = os.path.join(out_folder, f"case_{predict_case}_with_rnn_preds.csv")
df_test_pred.to_csv(out_file, index=False) df_test_pred.to_csv(out_file, index=False)
print(f"\nSaved predictions to: {out_file}") print(f"\nSaved predictions to: {out_file}")
print("Test RMSE summary:", rmse_by_target)
del final_model, test_loader, test_ds
gc.collect()
if torch.cuda.is_available():
torch.cuda.empty_cache()
if __name__ == "__main__": if __name__ == "__main__":
......
Code/src/hysteretic_curves/run.sh
#!/bin/bash #!/bin/bash
# python predict_hysteretic_curves.py --W 2 --B 29 --n-trials 25 python predict_hysteretic_curves.py --W 2 --B 29 --n-trials 25
python predict_hysteretic_curves.py --W 2 --B 34 --n-trials 25 python predict_hysteretic_curves.py --W 2 --B 34 --n-trials 25
python predict_hysteretic_curves.py --W 3 --B 29 --n-trials 25 python predict_hysteretic_curves.py --W 3 --B 29 --n-trials 25
python predict_hysteretic_curves.py --W 3 --B 34 --n-trials 25 python predict_hysteretic_curves.py --W 3 --B 34 --n-trials 25
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment