Upload 29 files

01_combining_dataframes_pipeline.py

@@ -0,0 +1,20 @@
+from sklearn.pipeline import Pipeline
+from pipeline_classes import ImportData, CreateCombinedDataFrame
+from _config import config
+import time
+
+# This pipeline combines the self-reports and accelerometer dataframes with a given timewindow into a single dataframe as a csv file
+combining_dataframes_pipeline = Pipeline([
+    ('import_data', ImportData(use_accel=True, use_reports=True, use_combined=False, use_features=False)),  # input path to self-reports data),
+    ('create_combined_dataframe', CreateCombinedDataFrame(time_window=config["time_window"], label_columns=config["label_columns"])),
+])
+
+# This will measure the time taken to run the pipeline
+start_time = time.time()
+
+# This will start the pipeline and return the combined dataframe
+output_df = combining_dataframes_pipeline.fit_transform(None)
+
+
+end_time = time.time()
+print(f"Time taken: {int((end_time - start_time) // 60)} minutes and {(end_time - start_time) % 60:.2f} seconds")

02_feature_extraction_pipeline.py

@@ -0,0 +1,27 @@
+from sklearn.pipeline import Pipeline
+from pipeline_classes import ImportData, LowPassFilter, ScaleXYZData, ExtractFeatures
+from _config import config
+import time
+
+# This pipeline extracts features from the combined dataframe and exports it to a csv file
+feature_extraction_pipeline = Pipeline([
+    ('import_data', ImportData(use_accel=False, use_reports=False, use_combined=True, use_features=False)), # input path to combined data
+    ('low_pass_filter', LowPassFilter(cutoff_frequency=config["cutoff_frequency"], sampling_rate=config["data_frequency"], order=config["order"])),
+    ('scale_xyz_data', ScaleXYZData(scaler_type=config["scaler_type"])),
+    ('extract_features', ExtractFeatures(window_length=config["window_length"],
+                                         window_step_size=config["window_step_size"],
+                                         data_frequency=config["data_frequency"],
+                                         selected_domains=config["selected_domains"],
+                                         include_magnitude=config["include_magnitude"],
+                                         label_columns=config["label_columns"])),
+])
+
+# This will measure the time taken to run the pipeline
+start_time = time.time()
+
+# This will start the pipeline and return the feature dataframe
+output_df = feature_extraction_pipeline.fit_transform(None)
+
+
+end_time = time.time()
+print(f"Time taken: {int((end_time - start_time) // 60)} minutes and {(end_time - start_time) % 60:.2f} seconds")

03_training_model_pipeline.py

@@ -0,0 +1,21 @@
+from sklearn.pipeline import Pipeline
+from pipeline_classes import ImportData, PCAHandler, TrainModel
+from _config import config
+import time
+
+# This pipeline trains a model on the feature dataframe and export the model to a pickle file and general information to a json file
+training_model_pipeline = Pipeline([
+    ('import_data', ImportData(use_accel=False, use_reports=False, use_combined=False, use_features=True)),
+    ('pca_handler', PCAHandler(apply_pca=config["apply_pca"], variance=config["pca_variance"])),
+    ('train_model', TrainModel(config=config)),
+])
+
+# This will measure the time taken to run the pipeline
+start_time = time.time()
+
+# This will start the pipeline and return the model and a report
+output_df = training_model_pipeline.fit_transform(None)
+
+
+end_time = time.time()
+print(f"Time taken: {int((end_time - start_time) // 60)} minutes and {(end_time - start_time) % 60:.2f} seconds")

04_analyzing_data_pipeline.py

@@ -0,0 +1,24 @@
+from sklearn.pipeline import Pipeline
+from pipeline_classes import ImportData, LowPassFilter, ScaleXYZData, ExtractFeatures, ClassifyMovementData
+from _config import config
+import time
+
+# This is the pipeline that will be used to analyze data which hasnt been classified yet and export the classified dataframe as a csv file
+analyzing_data_pipeline = Pipeline([
+    ('import_data', ImportData(use_accel=True, use_reports=False, use_combined=False, use_features=False)), # input path to accelerometer data)
+    ('low_pass_filter', LowPassFilter(cutoff_frequency=config["cutoff_frequency"], sampling_rate=config["data_frequency"], order=config["order"])),
+    ('scale_xyz_data', ScaleXYZData(scaler_type=config["scaler_type"])),
+    ('extract_features', ExtractFeatures(window_length=config['window_length'], window_step_size=config["window_step_size"], data_frequency=config["data_frequency"],
+                                          selected_domains=config['selected_domains'], include_magnitude=config['include_magnitude'])),
+    ('classify_movement_data', ClassifyMovementData()),
+])
+
+# This will measure the time taken to run the pipeline
+start_time = time.time()
+
+# This will start the pipeline and return the classified dataframe
+output_df = analyzing_data_pipeline.fit_transform(None)
+
+
+end_time = time.time()
+print(f"Time taken: {int((end_time - start_time) // 60)} minutes and {(end_time - start_time) % 60:.2f} seconds")

05_complete_trainmodel_pipeline.py

@@ -0,0 +1,30 @@
+from pipeline_classes import ImportData, LowPassFilter, ScaleXYZData, ExtractFeatures, CreateCombinedDataFrame, TrainModel, PCAHandler
+from _config import config
+from sklearn.pipeline import Pipeline
+import time
+
+# This is the complete pipeline that will be used to train a model on the combined dataframe and export the model to a pickle file and general information to a json file
+complete_training_model_pipeline = Pipeline([
+    ('import_data', ImportData(use_accel=True, use_reports=True, use_combined=False, use_features=False)),
+    ('create_combined_dataframe', CreateCombinedDataFrame(time_window=config["time_window"], label_columns=config["label_columns"])),
+    ('low_pass_filter', LowPassFilter(cutoff_frequency=config["cutoff_frequency"], sampling_rate=config["data_frequency"], order=config["order"])),
+    ('scale_xyz_data', ScaleXYZData(scaler_type=config["scaler_type"])),
+    ('extract_features', ExtractFeatures(window_length=config["window_length"],
+                                         window_step_size=config["window_step_size"],
+                                         data_frequency=config["data_frequency"],
+                                         selected_domains=config["selected_domains"],
+                                         include_magnitude=config["include_magnitude"],
+                                         label_columns=config["label_columns"])),
+    ('pca_handler', PCAHandler(apply_pca=config["apply_pca"], variance=config["pca_variance"])),
+    ('train_model', TrainModel(config=config)),
+])
+
+# This will measure the time taken to run the pipeline
+start_time = time.time()
+
+# This will start the pipeline and return the model and a report
+output_df = complete_training_model_pipeline.fit_transform(None)
+
+
+end_time = time.time()
+print(f"Time taken: {int((end_time - start_time) // 60)} minutes and {(end_time - start_time) % 60:.2f} seconds")

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 mininato
+
+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.

Selfreports.csv

@@ -0,0 +1,331 @@
+id,participantId,timeOfNotification,timeOfEngagement,timeOfContext,timeOfValence,timeOfArousal,valence,arousal,context
+1,1,459163,464533,497941,481778,491757,NEUTRAL,LOW,VIEW_BOOTH
+2,1,1116342,1123641,1147370,1137105,1142685,NEUTRAL,MEDIUM,CONVERSATION
+5,1,3075708,3082136,3115554,3105558,3109606,NEUTRAL,LOW,WALKING
+6,1,3562201,3582250,3603151,3593212,3597877,NEUTRAL,LOW,VIEW_BOOTH
+9,2,364259,373046,397051,380931,386833,NEUTRAL,MEDIUM,WALKING
+10,2,970402,983704,1000405,992294,996226,NEUTRAL,MEDIUM,CONVERSATION
+11,2,1610269,1626103,1644209,1634845,1639029,NEUTRAL,MEDIUM,WALKING
+15,3,418630,438014,474684,463238,467987,NEGATIVE,LOW,CONVERSATION
+16,3,1011885,1031630,1048935,1039714,1043750,NEGATIVE,LOW,CONVERSATION
+17,3,1764167,1768115,1794167,1777001,1790185,NEUTRAL,HIGH,CONVERSATION
+20,3,3518972,3522724,3555492,3541222,3551571,NEUTRAL,HIGH,CONVERSATION
+22,3,4852881,4929930,4974627,4963038,4970708,NEGATIVE,HIGH,CONVERSATION
+23,3,5685790,5689597,5717615,5708431,5713871,NEGATIVE,HIGH,CONVERSATION
+24,3,6169580,6191219,6207613,6200060,6203908,NEUTRAL,LOW,CONVERSATION
+26,3,7485712,7503454,7530876,7517871,7527148,NEGATIVE,LOW,CONVERSATION
+27,4,398322,418695,445787,432832,438441,NEUTRAL,MEDIUM,OTHER
+28,4,1207854,1213214,1234830,1225417,1231059,POSITIVE,HIGH,CONVERSATION
+29,4,1632001,1635069,1674502,1663489,1667863,NEGATIVE,LOW,CONVERSATION
+33,4,4078936,4082602,4110423,4099608,4106566,POSITIVE,MEDIUM,CONVERSATION
+34,4,4984332,4987509,5023409,5015376,5019515,NEUTRAL,LOW,CONVERSATION
+36,4,6258034,6306827,6350439,6341905,6346349,POSITIVE,LOW,CONVERSATION
+38,4,7537831,7560818,7584662,7576033,7579959,NEUTRAL,LOW,CONVERSATION
+42,5,1033450,1037216,1061281,1047854,1053884,NEUTRAL,MEDIUM,OTHER
+46,5,3810481,3814183,3832718,3824055,3828757,POSITIVE,MEDIUM,CONVERSATION
+49,6,996095,1003869,1040125,1024551,1031703,POSITIVE,MEDIUM,OTHER
+50,6,1984793,1987621,2006622,1996185,2001620,POSITIVE,MEDIUM,CONVERSATION
+51,6,2223253,2241585,2264887,2252721,2259434,POSITIVE,LOW,WALKING
+52,6,3257255,3285071,3302306,3294247,3298276,POSITIVE,MEDIUM,WALKING
+53,6,3482852,3486170,3503682,3494330,3499793,POSITIVE,MEDIUM,WALKING
+54,7,438176,457822,483704,468378,479593,NEUTRAL,MEDIUM,WALKING
+56,7,1641094,1656884,1677728,1669702,1673799,POSITIVE,MEDIUM,CONVERSATION
+58,7,2946746,2955282,2980581,2966335,2975176,NEUTRAL,LOW,VIEW_BOOTH
+59,7,3743733,3748179,3766580,3758612,3762573,NEUTRAL,MEDIUM,WALKING
+62,8,389047,420625,463986,436690,459735,POSITIVE,MEDIUM,CONVERSATION
+63,8,1004645,1008855,1040595,1023803,1032378,POSITIVE,MEDIUM,OTHER
+64,8,1919221,1931695,1992970,1969430,1988013,NEUTRAL,MEDIUM,CONVERSATION
+65,8,2304132,2321452,2351306,2331563,2341606,POSITIVE,MEDIUM,OTHER
+66,8,3121983,3131925,3486662,3157033,3482671,POSITIVE,MEDIUM,WALKING
+67,8,3516191,3519846,3541834,3529143,3536563,NEUTRAL,MEDIUM,VIEW_BOOTH
+68,9,440526,445315,465909,455321,459339,POSITIVE,HIGH,VIEW_BOOTH
+69,9,1054085,1057909,1080160,1070152,1075098,POSITIVE,HIGH,CONVERSATION
+71,9,2254651,2271819,3737241,2283233,2287237,NEUTRAL,HIGH,VIEW_BOOTH
+74,10,396440,414965,433816,423861,429852,NEUTRAL,LOW,CONVERSATION
+75,10,973814,991385,1015126,1005754,1010567,NEUTRAL,LOW,OTHER
+76,10,1714211,1718463,1744852,1732967,1739303,NEUTRAL,MEDIUM,VIEW_BOOTH
+77,10,2278818,2288071,2316326,2302447,2310891,NEUTRAL,MEDIUM,WALKING
+79,10,3485331,3502674,3522581,3514834,3518716,POSITIVE,HIGH,VIEW_BOOTH
+80,10,4139053,4143444,4183361,4172053,4178035,POSITIVE,MEDIUM,VIEW_BOOTH
+81,10,4728105,4731602,4748899,4740655,4744889,NEUTRAL,LOW,VIEW_BOOTH
+82,10,5417150,5432962,5451095,5442474,5446660,NEGATIVE,LOW,OTHER
+83,10,5942145,5951917,5968416,5960758,5964570,POSITIVE,MEDIUM,OTHER
+84,10,6725895,6737449,6753352,6745452,6749503,NEUTRAL,MEDIUM,OTHER
+85,10,7191467,7195461,7211190,7203423,7207195,NEGATIVE,LOW,OTHER
+86,10,7952233,7968095,7984532,7976096,7980614,NEUTRAL,MEDIUM,OTHER
+87,10,8492220,8499887,8522413,8514738,8518572,NEUTRAL,MEDIUM,OTHER
+88,10,9170759,9187921,9212255,9203547,9207313,NEUTRAL,MEDIUM,WALKING
+89,11,360955,389975,412833,398929,403465,POSITIVE,LOW,OTHER
+90,11,972477,977166,997502,985754,989832,NEUTRAL,LOW,OTHER
+91,11,1623997,1630973,1647457,1639451,1643935,NEUTRAL,MEDIUM,OTHER
+92,11,2275129,2280181,2296938,2288590,2292975,NEUTRAL,LOW,OTHER
+93,11,2911933,2918540,2934638,2926380,2930641,NEUTRAL,LOW,OTHER
+94,11,3538928,3544360,3560998,3553040,3557121,NEUTRAL,MEDIUM,OTHER
+95,12,471215,528562,557396,544878,551091,NEGATIVE,LOW,OTHER
+96,12,1225258,1280326,1299757,1292099,1296126,POSITIVE,MEDIUM,OTHER
+97,12,1836039,1840018,1861866,1850346,1857708,POSITIVE,LOW,OTHER
+98,12,2486363,2492365,2512306,2503087,2507925,NEUTRAL,LOW,OTHER
+99,12,3137287,3145520,3164396,3153542,3160267,POSITIVE,LOW,OTHER
+100,12,3789210,3796685,3816312,3807352,3811878,NEGATIVE,LOW,OTHER
+101,12,4440308,4510695,4535084,4524621,4531188,NEGATIVE,HIGH,WALKING
+106,13,2224206,2240747,2256354,2248791,2252390,NEUTRAL,HIGH,WALKING
+111,14,403464,420660,440265,429027,431958,POSITIVE,MEDIUM,VIEW_BOOTH
+112,14,969390,979347,998577,989891,992501,NEUTRAL,MEDIUM,VIEW_BOOTH
+113,14,1682520,1698696,1712380,1706198,1709353,NEUTRAL,HIGH,OTHER
+114,14,2210289,2218612,2236363,2231282,2233720,NEUTRAL,LOW,OTHER
+115,14,2899497,2904801,2918423,2911879,2916203,NEUTRAL,MEDIUM,OTHER
+116,14,3484157,3488140,3496411,3492407,3494872,NEUTRAL,MEDIUM,WALKING
+117,14,4221518,4225648,4241502,4237724,4239472,NEUTRAL,MEDIUM,VIEW_BOOTH
+118,14,4746426,4755810,4792302,4784322,4787508,NEUTRAL,MEDIUM,VIEW_BOOTH
+119,14,5422344,5460655,5468614,5464395,5466205,POSITIVE,HIGH,OTHER
+120,14,6037674,6058817,6066762,6062339,6064123,POSITIVE,HIGH,OTHER
+121,14,6638455,6647724,6655649,6651931,6653654,POSITIVE,MEDIUM,OTHER
+122,15,437376,441581,472237,454875,467599,NEUTRAL,MEDIUM,CONVERSATION
+123,15,992469,996800,1027268,1012815,1023064,NEUTRAL,MEDIUM,WALKING
+128,15,4195442,4199293,4216706,4208418,4212390,POSITIVE,HIGH,OTHER
+129,15,4833093,4853849,4869948,4862464,4866272,POSITIVE,MEDIUM,VIEW_BOOTH
+130,15,5704140,5709092,5726903,5717912,5722864,NEUTRAL,MEDIUM,VIEW_BOOTH
+131,15,6065767,6069742,6092620,6082637,6088334,NEUTRAL,LOW,VIEW_BOOTH
+132,15,7212321,7217115,7235556,7227281,7231793,NEUTRAL,MEDIUM,VIEW_BOOTH
+133,15,7322566,7326529,7370005,7336440,7340684,NEUTRAL,MEDIUM,OTHER
+135,15,8649125,8656903,8675123,8665598,8670989,NEUTRAL,MEDIUM,OTHER
+137,16,1032124,1044697,1087131,1058687,1070082,NEUTRAL,HIGH,VIEW_BOOTH
+138,16,1656017,1660069,1686740,1674587,1678965,POSITIVE,MEDIUM,VIEW_BOOTH
+139,16,2346461,2366522,2392805,2383298,2388594,POSITIVE,LOW,VIEW_BOOTH
+140,16,2864762,2875565,2901365,2889232,2895220,NEUTRAL,MEDIUM,VIEW_BOOTH
+141,16,3580123,3587169,3607438,3597425,3602877,POSITIVE,LOW,VIEW_BOOTH
+143,17,1177480,1187827,1207407,1197959,1202464,POSITIVE,MEDIUM,CONVERSATION
+145,17,2378390,2395909,2414759,2405054,2409389,POSITIVE,HIGH,OTHER
+148,17,4127579,4132459,4154750,4144095,4149272,NEUTRAL,HIGH,OTHER
+150,17,5372478,5379120,5398635,5389985,5394418,POSITIVE,MEDIUM,OTHER
+157,18,3190342,3215151,3242119,3226979,3235162,POSITIVE,MEDIUM,CONVERSATION
+161,19,1294141,1415217,1431226,1424046,1426928,POSITIVE,MEDIUM,CONVERSATION
+162,19,1955322,1961892,1972531,1966683,1969333,POSITIVE,HIGH,CONVERSATION
+163,19,2601157,2629622,2651502,2641481,2646733,POSITIVE,MEDIUM,OTHER
+164,19,3164532,3169232,3185048,3175434,3178806,NEUTRAL,MEDIUM,CONVERSATION
+165,19,3921753,3925215,3939677,3931322,3934806,NEUTRAL,HIGH,WALKING
+166,20,373771,386930,419180,397258,410188,POSITIVE,MEDIUM,VIEW_BOOTH
+170,20,3165605,3170422,3202474,3188909,3196662,NEUTRAL,MEDIUM,OTHER
+171,20,3805263,3810431,3830273,3820936,3826839,NEUTRAL,MEDIUM,OTHER
+173,20,6758600,6763765,6784057,6775562,6780228,NEGATIVE,MEDIUM,OTHER
+174,21,385802,393372,410816,401501,405852,POSITIVE,HIGH,CONVERSATION
+177,21,2532744,2539297,2561349,2550342,2554795,NEUTRAL,MEDIUM,VIEW_BOOTH
+178,21,3172740,3178901,3198546,3190166,3194151,NEUTRAL,MEDIUM,OTHER
+179,21,3827443,3834605,3853997,3844661,3849645,NEUTRAL,LOW,OTHER
+180,21,4479785,4486212,4504725,4495182,4499348,NEUTRAL,LOW,OTHER
+181,21,5131181,5139047,5159896,5146523,5150753,POSITIVE,MEDIUM,CONVERSATION
+182,21,5711830,5714978,5737651,5729696,5733514,NEUTRAL,HIGH,CONVERSATION
+183,21,6683983,6692843,6710683,6702965,6706728,POSITIVE,HIGH,OTHER
+184,21,7019082,7026023,7048869,7035489,7045423,NEUTRAL,MEDIUM,CONVERSATION
+185,22,428119,434457,454721,442558,446783,NEUTRAL,MEDIUM,OTHER
+186,22,991572,1000928,1016396,1008673,1012737,NEUTRAL,LOW,OTHER
+187,22,1669957,1673695,1689053,1680980,1684893,NEUTRAL,LOW,OTHER
+189,22,3049656,3054862,3071589,3063222,3067288,NEUTRAL,MEDIUM,OTHER
+190,22,3477857,3490396,3505673,3498249,3502057,NEUTRAL,LOW,OTHER
+191,22,4315873,4320309,4335554,4327698,4331441,NEGATIVE,MEDIUM,OTHER
+192,22,4722764,4730027,4746527,4738793,4742585,NEGATIVE,LOW,OTHER
+193,22,5548660,5563012,5578725,5570634,5574928,NEUTRAL,MEDIUM,OTHER
+194,22,5941373,5947167,5962518,5954889,5958740,NEUTRAL,MEDIUM,OTHER
+195,22,6851905,6856347,6875883,6867202,6871749,NEGATIVE,LOW,OTHER
+196,22,7186700,7191185,7206608,7198813,7202627,NEUTRAL,MEDIUM,OTHER
+198,23,1047926,1055012,1102461,1071688,1083843,NEUTRAL,LOW,CONVERSATION
+199,23,1772464,1792462,1835713,1804059,1832868,NEUTRAL,MEDIUM,CONVERSATION
+203,23,4547430,4586395,4599023,4591266,4595541,POSITIVE,HIGH,CONVERSATION
+204,23,4974672,4979399,4997970,4988056,4995167,POSITIVE,HIGH,WALKING
+205,23,5817273,5821397,5836072,5829665,5834083,NEUTRAL,HIGH,CONVERSATION
+207,23,7031913,7049611,7084672,7071116,7074149,POSITIVE,HIGH,CONVERSATION
+208,23,7550340,7597681,7607995,7603163,7606027,NEUTRAL,MEDIUM,CONVERSATION
+209,23,8409647,8415493,8437908,8424950,8427100,POSITIVE,MEDIUM,WALKING
+210,24,375235,379813,393686,386003,389556,NEUTRAL,MEDIUM,WALKING
+211,24,972962,976795,991889,984019,986248,POSITIVE,MEDIUM,OTHER
+212,24,1591410,1612868,1622383,1618687,1620474,POSITIVE,MEDIUM,OTHER
+213,24,2200900,2208166,2219359,2213865,2217156,POSITIVE,MEDIUM,OTHER
+214,24,2857829,2860854,2869621,2865236,2866957,NEUTRAL,MEDIUM,CONVERSATION
+215,24,3457354,3473420,3483402,3479330,3481058,NEUTRAL,MEDIUM,OTHER
+216,24,4071613,4085051,4094890,4090372,4093367,POSITIVE,LOW,VIEW_BOOTH
+217,24,4767647,4771718,4782830,4777874,4780207,NEUTRAL,LOW,CONVERSATION
+218,24,5385676,5432938,5443650,5439714,5441897,POSITIVE,MEDIUM,VIEW_BOOTH
+220,24,6628946,6632140,6643607,6639370,6641237,POSITIVE,LOW,CONVERSATION
+221,24,7474159,7489732,7496711,7493660,7494998,POSITIVE,HIGH,CONVERSATION
+222,24,8131082,8135229,8142779,8139652,8141347,POSITIVE,MEDIUM,VIEW_BOOTH
+223,24,8687266,8867182,8892807,8889186,8890950,NEUTRAL,LOW,VIEW_BOOTH
+224,25,374839,415756,436542,427045,432265,POSITIVE,HIGH,WALKING
+225,25,1158058,1162323,1171044,1166086,1168793,NEUTRAL,HIGH,CONVERSATION
+226,25,1783694,1790040,1800330,1795452,1797988,POSITIVE,MEDIUM,WALKING
+227,25,2456368,2466635,2473758,2470022,2471922,POSITIVE,MEDIUM,OTHER
+228,25,3016822,3024054,3031747,3027717,3029548,POSITIVE,MEDIUM,OTHER
+229,25,3705334,3720473,3728658,3724873,3726649,NEUTRAL,MEDIUM,OTHER
+230,25,4236064,4261939,4268459,4265131,4266954,NEUTRAL,MEDIUM,CONVERSATION
+231,25,5139002,5142634,5152593,5148598,5150577,NEGATIVE,MEDIUM,VIEW_BOOTH
+232,25,5439118,5516441,5522658,5519443,5521127,POSITIVE,HIGH,CONVERSATION
+233,25,6438209,6457298,6465349,6460240,6463427,NEGATIVE,LOW,CONVERSATION
+234,25,6739006,6803282,6812301,6806965,6808351,NEGATIVE,LOW,OTHER
+237,26,1004099,1011612,1028117,1020592,1023931,POSITIVE,MEDIUM,WALKING
+239,26,2232180,2250890,2260873,2255161,2257449,POSITIVE,MEDIUM,OTHER
+241,26,3438317,3442612,3465619,3447877,3454150,NEUTRAL,MEDIUM,OTHER
+245,26,5960595,5964255,5977394,5973014,5975368,POSITIVE,HIGH,CONVERSATION
+246,27,536701,543426,572213,559651,565921,POSITIVE,MEDIUM,OTHER
+247,27,1189638,1194293,1211103,1203143,1205334,NEUTRAL,MEDIUM,OTHER
+248,27,1974812,1979228,1995853,1991659,1993797,NEUTRAL,MEDIUM,OTHER
+249,27,2467550,2492439,2508884,2501964,2507068,NEUTRAL,LOW,OTHER
+250,27,3196398,3200121,3210968,3204262,3208914,NEUTRAL,LOW,OTHER
+251,27,3677141,3682487,3698420,3694521,3696439,POSITIVE,MEDIUM,WALKING
+252,27,4737742,4743456,4765589,4759698,4763171,NEUTRAL,LOW,VIEW_BOOTH
+253,27,4892215,4935051,4944043,4939719,4942123,NEUTRAL,MEDIUM,CONVERSATION
+255,27,6492416,6496291,6511948,6505695,6509294,NEUTRAL,LOW,WALKING
+257,27,7847871,7908945,7920353,7916658,7918473,NEGATIVE,LOW,CONVERSATION
+260,28,476606,482399,495293,489895,492119,POSITIVE,MEDIUM,CONVERSATION
+261,28,1001075,1006696,1022210,1013181,1018261,NEGATIVE,HIGH,OTHER
+262,28,1799545,1805471,1814420,1810339,1812250,NEGATIVE,MEDIUM,WALKING
+263,28,2334532,2339044,2350019,2343825,2345735,NEUTRAL,MEDIUM,CONVERSATION
+264,28,3047103,3057355,3064855,3060992,3062786,NEUTRAL,MEDIUM,CONVERSATION
+265,28,3643104,3660732,3669348,3665989,3667746,POSITIVE,HIGH,CONVERSATION
+267,28,5114105,5121522,5131232,5126904,5129271,NEUTRAL,MEDIUM,WALKING
+268,28,5572967,5590179,5596978,5593618,5595145,NEUTRAL,MEDIUM,WALKING
+270,28,6996539,7028566,7037503,7032028,7034129,POSITIVE,HIGH,VIEW_BOOTH
+273,29,458619,466826,485566,475687,478805,POSITIVE,MEDIUM,WALKING
+274,29,1018280,1022240,1045166,1040684,1042630,POSITIVE,HIGH,CONVERSATION
+275,29,1896030,1900823,1911410,1906091,1909245,NEUTRAL,HIGH,VIEW_BOOTH
+277,29,3168088,3180806,3191350,3185135,3188772,NEGATIVE,LOW,VIEW_BOOTH
+278,29,3799169,3963759,3979611,3970320,3974500,POSITIVE,MEDIUM,CONVERSATION
+279,29,4595223,4599365,4610211,4604164,4607831,NEUTRAL,MEDIUM,CONVERSATION
+280,29,5225253,5227892,5236898,5232521,5234335,NEUTRAL,MEDIUM,VIEW_BOOTH
+282,29,6444878,6447815,6699311,6452365,6455099,POSITIVE,MEDIUM,OTHER
+283,30,447369,466334,486285,476298,481761,NEUTRAL,LOW,WALKING
+284,30,1014547,1018455,1030259,1023458,1026962,POSITIVE,HIGH,OTHER
+285,30,1813364,1816159,1830888,1824864,1827496,NEUTRAL,LOW,CONVERSATION
+286,30,2225313,2229694,2236292,2233304,2234671,POSITIVE,HIGH,CONVERSATION
+287,30,3123141,3127754,3135059,3131141,3133218,NEUTRAL,LOW,CONVERSATION
+288,30,3469211,3472891,3480588,3476766,3478485,POSITIVE,HIGH,CONVERSATION
+289,30,4447626,4450357,4458639,4455220,4456848,POSITIVE,MEDIUM,CONVERSATION
+290,30,4706883,4712173,4719855,4715992,4717904,POSITIVE,MEDIUM,CONVERSATION
+291,30,5668802,5673738,5686133,5680486,5682460,POSITIVE,MEDIUM,CONVERSATION
+292,30,5909192,5912558,5920920,5916664,5918603,NEUTRAL,MEDIUM,OTHER
+293,30,6894924,6899530,6908985,6903811,6905595,NEGATIVE,MEDIUM,VIEW_BOOTH
+294,30,7206549,7210380,7218136,7214044,7216234,POSITIVE,MEDIUM,CONVERSATION
+295,30,8175968,8188976,8197301,8194089,8195626,NEUTRAL,MEDIUM,CONVERSATION
+296,31,450055,456852,470741,464492,467490,POSITIVE,MEDIUM,WALKING
+297,31,1027972,1031999,1041707,1036536,1039626,NEUTRAL,MEDIUM,CONVERSATION
+298,31,1864189,1870609,1880504,1875238,1878668,POSITIVE,MEDIUM,VIEW_BOOTH
+299,31,2524865,2594119,2601479,2598005,2599792,POSITIVE,HIGH,VIEW_BOOTH
+300,31,3095696,3108949,3118228,3113028,3115879,NEUTRAL,MEDIUM,CONVERSATION
+301,31,3957981,3972387,3979397,3976233,3977939,POSITIVE,HIGH,VIEW_BOOTH
+302,31,4585810,4627485,4634986,4630965,4632966,NEUTRAL,MEDIUM,WALKING
+303,31,5232846,5241562,5257078,5245032,5255390,POSITIVE,HIGH,VIEW_BOOTH
+304,31,5806903,5852567,5861047,5856299,5859070,NEUTRAL,MEDIUM,CONVERSATION
+305,31,6473876,6491700,6499561,6495728,6497253,POSITIVE,MEDIUM,VIEW_BOOTH
+306,32,415875,465231,490166,482402,485624,NEUTRAL,MEDIUM,WALKING
+307,32,1019863,1025362,1045056,1030184,1035107,POSITIVE,HIGH,OTHER
+308,32,1815296,1817977,1833558,1823311,1829782,NEUTRAL,MEDIUM,CONVERSATION
+309,32,2303442,2333455,2345887,2340387,2343551,NEUTRAL,MEDIUM,VIEW_BOOTH
+310,32,3480594,4767318,4775586,4772166,4773762,NEUTRAL,MEDIUM,CONVERSATION
+313,32,4930815,4940270,4949457,4944587,4946735,POSITIVE,HIGH,WALKING
+314,32,5970857,5978016,5991748,5986972,5989422,NEGATIVE,MEDIUM,VIEW_BOOTH
+315,32,6164148,6171884,6180697,6176540,6178445,NEUTRAL,MEDIUM,WALKING
+317,32,7406065,7412599,7423023,7418435,7420389,NEUTRAL,MEDIUM,OTHER
+319,33,413389,508962,521373,514056,516476,POSITIVE,MEDIUM,WALKING
+320,33,1064795,1073304,1088584,1083644,1085745,POSITIVE,MEDIUM,VIEW_BOOTH
+321,33,1702340,1735244,1744899,1739908,1742719,POSITIVE,HIGH,VIEW_BOOTH
+322,33,2354336,2426868,2435258,2430689,2433579,POSITIVE,HIGH,VIEW_BOOTH
+323,33,2925085,2945096,2952927,2948622,2951113,POSITIVE,HIGH,WALKING
+324,33,3568527,3713623,3721992,3717477,3719842,POSITIVE,HIGH,VIEW_BOOTH
+328,33,6078250,6175510,6185420,6179357,6182878,POSITIVE,HIGH,VIEW_BOOTH
+330,33,7428337,7546845,7554072,7550332,7552007,POSITIVE,HIGH,CONVERSATION
+331,33,8042241,8096825,8103922,8100257,8102092,POSITIVE,HIGH,CONVERSATION
+332,33,8755173,8778412,8786296,8781974,8783745,POSITIVE,HIGH,OTHER
+333,34,447403,599386,611664,605102,607625,POSITIVE,HIGH,CONVERSATION
+334,34,990543,1007641,1018160,1013479,1016629,NEUTRAL,MEDIUM,WALKING
+335,34,1754661,1759063,1769011,1764134,1765991,NEUTRAL,MEDIUM,CONVERSATION
+336,34,2376149,2380973,2393972,2386984,2391550,NEUTRAL,HIGH,WALKING
+338,34,3691582,3695655,3709449,3703057,3705195,NEUTRAL,LOW,OTHER
+340,35,469206,487847,516695,500694,503836,POSITIVE,MEDIUM,OTHER
+341,35,1060038,1065202,1082202,1074460,1077200,NEUTRAL,MEDIUM,WALKING
+342,35,1724995,1729403,1750704,1742459,1745993,NEUTRAL,MEDIUM,CONVERSATION
+344,35,2937105,2942275,2956687,2949676,2951460,NEUTRAL,MEDIUM,CONVERSATION
+346,35,4209228,4212499,4221547,4216521,4218310,NEUTRAL,MEDIUM,CONVERSATION
+347,35,4971865,4979279,4989400,4984380,4986488,NEUTRAL,MEDIUM,CONVERSATION
+348,35,5518036,5521107,5531601,5525542,5527340,NEUTRAL,MEDIUM,WALKING
+349,35,6228270,6244592,6253785,6250388,6251858,NEUTRAL,MEDIUM,CONVERSATION
+350,35,6798570,6806229,6814386,6810912,6812601,NEUTRAL,MEDIUM,WALKING
+352,35,8075559,8093519,8100809,8096905,8098683,NEUTRAL,MEDIUM,CONVERSATION
+353,35,8769411,8772863,8782066,8777050,8778714,NEUTRAL,MEDIUM,OTHER
+354,36,447096,462169,473327,468716,470877,POSITIVE,HIGH,WALKING
+355,36,1013861,1019472,1028712,1023114,1025523,POSITIVE,HIGH,OTHER
+356,36,1807896,1813955,1821801,1817652,1819757,NEUTRAL,MEDIUM,WALKING
+357,36,2220363,2224940,2232741,2228898,2230450,POSITIVE,HIGH,VIEW_BOOTH
+358,36,3101101,3104983,3111671,3108253,3110076,POSITIVE,HIGH,VIEW_BOOTH
+359,36,3512204,3528567,3535243,3532199,3533668,POSITIVE,HIGH,VIEW_BOOTH
+360,36,4416861,4433129,4439381,4436250,4437921,POSITIVE,HIGH,VIEW_BOOTH
+362,36,5645858,5664499,5670490,5667661,5669057,POSITIVE,HIGH,VIEW_BOOTH
+363,36,6164371,6169127,6175286,6172464,6173816,NEUTRAL,MEDIUM,WALKING
+365,36,7421739,7425039,7434355,7430565,7432070,POSITIVE,HIGH,WALKING
+366,36,8227815,8230769,8237425,8233706,8235518,NEUTRAL,HIGH,WALKING
+367,37,403144,408364,427858,420480,424330,POSITIVE,HIGH,WALKING
+368,37,1012406,1165451,1184402,1170892,1172800,POSITIVE,HIGH,OTHER
+369,37,1637832,1655297,1667069,1662090,1664478,NEUTRAL,MEDIUM,OTHER
+370,37,2247122,2258451,2272945,2269396,2271412,NEGATIVE,LOW,OTHER
+371,37,3055252,3058800,3068584,3064767,3066722,NEUTRAL,MEDIUM,OTHER
+372,37,3460162,3475885,3491098,3481140,3483810,NEGATIVE,LOW,OTHER
+374,37,4692798,4695962,4740135,4705077,4724049,POSITIVE,HIGH,CONVERSATION
+375,37,5596301,5613971,5641687,5621133,5639784,NEUTRAL,LOW,WALKING
+377,37,6892400,6898018,6920070,6903743,6912592,POSITIVE,MEDIUM,CONVERSATION
+378,37,7490362,7493636,7517280,7511366,7513696,POSITIVE,MEDIUM,CONVERSATION
+379,37,8142726,8146314,8155828,8150876,8152986,NEUTRAL,LOW,VIEW_BOOTH
+380,37,8721545,8861944,8917474,8913390,8915337,POSITIVE,HIGH,CONVERSATION
+381,38,379560,428257,453470,437070,447170,POSITIVE,MEDIUM,WALKING
+386,38,3439806,3463018,3473817,3468198,3470588,NEUTRAL,MEDIUM,CONVERSATION
+388,38,4669425,4697990,4708408,4701833,4704267,NEUTRAL,MEDIUM,VIEW_BOOTH
+389,38,5470509,5482379,5496805,5492934,5494724,POSITIVE,LOW,OTHER
+390,38,5984006,6025754,6039144,6035022,6037323,POSITIVE,MEDIUM,WALKING
+391,38,6720527,6757685,6768349,6764563,6766228,NEUTRAL,MEDIUM,OTHER
+392,38,7322979,7328006,7334304,7331241,7332799,NEUTRAL,MEDIUM,WALKING
+393,39,411121,416854,439176,431119,434850,NEUTRAL,MEDIUM,WALKING
+394,39,1018886,1031963,1047671,1037559,1042847,POSITIVE,HIGH,CONVERSATION
+395,39,1682708,1692246,1701295,1697346,1699068,POSITIVE,HIGH,CONVERSATION
+396,39,2226000,2233289,2248330,2239088,2240988,NEUTRAL,MEDIUM,OTHER
+397,39,2898988,2904659,2913307,2908750,2910443,POSITIVE,HIGH,OTHER
+398,39,3442347,3461449,3482733,3473144,3475355,NEUTRAL,LOW,OTHER
+399,39,4156358,4185857,4197449,4193803,4195452,NEUTRAL,MEDIUM,CONVERSATION
+400,39,4686398,4702686,4717298,4710614,4712251,NEUTRAL,MEDIUM,OTHER
+401,39,5419179,5440077,5448412,5444012,5446784,POSITIVE,HIGH,CONVERSATION
+402,39,5971099,5975971,5983904,5979615,5982259,POSITIVE,MEDIUM,CONVERSATION
+403,39,6686024,6694008,6702015,6698583,6700281,NEUTRAL,MEDIUM,CONVERSATION
+404,39,7249240,7253659,7266689,7262878,7264509,NEUTRAL,MEDIUM,CONVERSATION
+405,40,385066,418772,438171,426505,432094,POSITIVE,MEDIUM,CONVERSATION
+406,40,1046847,1188385,1200383,1192883,1195740,POSITIVE,MEDIUM,CONVERSATION
+409,40,3252870,3259774,3272231,3266838,3270230,POSITIVE,MEDIUM,CONVERSATION
+410,40,3625362,3628713,3642191,3637443,3639543,NEUTRAL,MEDIUM,CONVERSATION
+411,40,4498835,4531239,4542755,4537150,4539153,POSITIVE,HIGH,WALKING
+412,40,4837473,4892195,4907013,4897288,4904155,POSITIVE,HIGH,VIEW_BOOTH
+413,40,6240001,6288875,6297303,6292953,6295268,POSITIVE,MEDIUM,VIEW_BOOTH
+414,40,6240028,6271951,6285074,6278257,6281971,POSITIVE,MEDIUM,VIEW_BOOTH
+416,40,7502843,7510644,7521914,7517482,7519338,POSITIVE,LOW,OTHER
+419,41,415228,538718,569297,551726,556277,POSITIVE,MEDIUM,VIEW_BOOTH
+420,41,979467,983049,994709,989715,992779,POSITIVE,HIGH,CONVERSATION
+421,41,1675554,1782475,1806678,1794142,1802146,NEUTRAL,MEDIUM,CONVERSATION
+422,41,2187331,2192311,2201057,2197583,2199180,POSITIVE,HIGH,CONVERSATION
+423,41,2968943,2972945,3053893,3016814,3041099,POSITIVE,LOW,CONVERSATION
+424,41,3452560,3470742,3480266,3475005,3477156,POSITIVE,MEDIUM,OTHER
+425,41,4175648,4180301,4194774,4186448,4191673,POSITIVE,LOW,CONVERSATION
+426,41,4832504,5004090,5014444,5009923,5011618,NEUTRAL,LOW,VIEW_BOOTH
+427,41,5384250,5466995,5474403,5471036,5472840,POSITIVE,MEDIUM,CONVERSATION
+428,41,6096933,6126741,6144612,6137039,6142446,POSITIVE,MEDIUM,CONVERSATION
+429,42,384228,399210,412073,405380,407533,NEUTRAL,MEDIUM,WALKING
+432,42,2207313,2212717,2230043,2222234,2223985,NEGATIVE,MEDIUM,WALKING
+434,42,3412215,3415935,3423645,3419517,3421448,POSITIVE,MEDIUM,CONVERSATION
+440,43,405159,413338,430651,419832,423368,POSITIVE,MEDIUM,WALKING
+443,43,2438678,2471892,2483633,2478961,2481228,POSITIVE,MEDIUM,CONVERSATION
+445,43,3875682,3883277,3902504,3888155,3891527,NEUTRAL,MEDIUM,VIEW_BOOTH
+446,43,4355656,4381506,4391650,4385945,4389476,POSITIVE,HIGH,CONVERSATION
+447,43,5145408,5151287,5166824,5159387,5163519,NEUTRAL,MEDIUM,CONVERSATION
+448,43,5805735,5812887,5821945,5816274,5820423,POSITIVE,MEDIUM,CONVERSATION
+449,43,6383233,6411042,6419546,6415165,6417611,NEUTRAL,MEDIUM,CONVERSATION
+451,43,7624508,7629824,7647099,7636938,7640844,NEUTRAL,LOW,WALKING
+452,43,8400302,8414184,8422698,8418167,8420180,POSITIVE,MEDIUM,WALKING
+453,44,455877,459825,473866,466570,469765,POSITIVE,HIGH,CONVERSATION
+455,44,1856736,1862210,1875706,1869148,1871332,NEGATIVE,MEDIUM,WALKING
+456,44,2480575,2592920,2602657,2596802,2598906,POSITIVE,MEDIUM,CONVERSATION
+457,44,3149594,3163102,3179345,3174258,3176756,POSITIVE,MEDIUM,VIEW_BOOTH
+460,44,4929682,4937300,4951621,4947538,4949290,NEUTRAL,MEDIUM,VIEW_BOOTH
+461,44,5666736,5670460,5688609,5676481,5678315,NEUTRAL,MEDIUM,VIEW_BOOTH
+462,44,6163589,6182378,6209246,6186723,6201830,POSITIVE,HIGH,VIEW_BOOTH

_config.py

@@ -0,0 +1,57 @@
+# Configuration file for the pipeline
+
+config = {
+    # Paths for Import Data
+    "accel_path": "Path or Name of File: ",  # Path to the accelerometer data
+    "reports_path": "Path or Name of SelfReports File",  # Path to the self-reports data
+    "combined_data_path": "Path or Name of File of Combined Data File",  # Path to the combined data
+    "features_data_path": "Path or Name of File of Features Data File",  # Path to the features data
+    "model_path": "Path or Name of Trained Model File",  # Path to the trained model
+
+    # Label Configuration
+    "label_columns": ["valence", "arousal"],  # Here you should input the emotion-labels that you are using
+    "target_label": "arousal",  # This is the target label that you want to predict (Only one label can be selected)
+
+    # Configuration for combined data
+    "time_window": 2,  # Minutes before and after the self-report
+
+    # Configuration for feature extraction
+    "window_length": 60,  # Window length in seconds / 60
+    "window_step_size": 20,  # Step size in seconds / 10%-50% of window_length / 20
+    "data_frequency": 25,  # Data frequency in Hz
+    "selected_domains": None,  # Default: Every domain / 'time_domain', 'spatial', 'frequency', 'statistical', 'wavelet' / multiple domains: ["time_domain", "frequency"] / order is not important
+    "include_magnitude": True,  # Include magnitude-based features or not
+
+    #Configuration for Low-pass filter
+    "cutoff_frequency": 10,  # Cut-off frequency for the low-pass filter
+    "order": 4,  # Order of the filter
+
+    # Configuration for Scaling
+    "scaler_type": "standard",  # Possible Scaler: 'standard' or 'minmax'
+
+    # Configuration for PCA
+    "apply_pca": False,  # Apply PCA or not
+    "pca_variance": 0.95,  # PCA variance threshold
+
+    # Configuration for model training
+    "classifier": "xgboost",  # Default classifier ('xgboost', 'svm', 'randomforest')
+
+    # Configuration for hyperparameter tuning
+    "n_splits": 5, # Number of splits for cross-validation
+    "n_iter": 30,   # Number of iterations for hyperparameter tuning
+    "n_jobs": -1,   # Number of jobs for parallel processing
+    "n_points": 1,  # Number of points to sample in the hyperparameter space
+
+    # If users want to define custom param_space, they can specify it here
+    "param_space": {
+        "learning_rate": (0.05, 0.2), 
+        "n_estimators": (200, 800),
+        "max_depth": (4, 8),
+        "min_child_weight": (1, 5),
+        "subsample": (0.6, 0.9),
+        "colsample_bytree": (0.6, 0.9),
+        "gamma": (0, 5),
+        "reg_alpha": (0, 5),
+        "reg_lambda": (0, 5)
+    },  # Set to {None} to use default inside the TrainModel class
+}

pipeline_classes/__init__.py

@@ -0,0 +1,10 @@
+# Description: This file is used to import all the classes in the pipeline_classes folder.
+
+from .import_data import ImportData
+from .create_combineddataframe import CreateCombinedDataFrame
+from .scale_xyzdata import ScaleXYZData
+from .extract_features import ExtractFeatures
+from .pcahandler import PCAHandler
+from .train_model import TrainModel
+from .classify_movementdata import ClassifyMovementData
+from .lowpassfilter import LowPassFilter

pipeline_classes/classify_movementdata.py

@@ -0,0 +1,35 @@
+import pandas as pd
+import numpy as np
+from sklearn.base import BaseEstimator, TransformerMixin
+import joblib
+from _config import config
+
+# This class is used to classify the movement data using a pre-trained model
+class ClassifyMovementData(BaseEstimator, TransformerMixin):
+    def __init__(self, model_path=None):
+        self.model_path = model_path if model_path else config.get("model_path")
+        self.model = None
+
+    def fit(self, X, y=None):
+        return self
+
+    def transform(self, X):
+        if self.model is None:
+            self.model = joblib.load(self.model_path)  # Load the pre-trained model
+            print(f"Model loaded from {self.model_path}")
+
+        # Assuming `X` is a DataFrame of pre-extracted features.
+        predictions = self.model.predict(X)
+
+        # Adding predictions to the DataFrame as the first column
+        X.insert(0, 'predicted_emotion', predictions)
+
+        print("Data classified successfully.")
+        
+        # Export the labeled DataFrame to CSV
+        window_length_str = str(config["window_length"])
+        output_file = f"classified_movement_data_window_{window_length_str}.csv"
+        X.to_csv(output_file, index=False)
+        print(f"Classified movement data exported successfully to {output_file}.")
+
+        return X 

pipeline_classes/create_combineddataframe.py

@@ -0,0 +1,83 @@
+import pandas as pd
+import numpy as np
+from sklearn.base import BaseEstimator, TransformerMixin
+from _config import config
+    
+class CreateCombinedDataFrame(BaseEstimator, TransformerMixin):
+    def __init__(self, time_window, label_columns=None):
+        self.time_window = time_window
+        self.label_columns = label_columns #if label_columns else ["arousal", "valence"]  # Default to arousal and valence if not specified
+
+    def fit(self, X, y=None):
+        return self
+    
+    def transform(self, X):
+        df_reports, df_accel = X
+
+        print(f"PreprocesssingCombined initialized with label_columns: {self.label_columns}")
+        # Ensure the chosen label columns exist in the dataset
+        valid_conditions = (df_reports['timeOfEngagement'] != 0) 
+        for label in self.label_columns:
+            valid_conditions &= (df_reports[label] != "NONE")
+        
+        df_reports = df_reports[valid_conditions].copy()
+
+        # Renaming and datetime conversion remains the same
+        df_accel.rename(columns={'timestamp': 'timeOfNotification'}, inplace=True)
+        df_accel["timeOfNotification"] = pd.to_datetime(df_accel["timeOfNotification"], unit="ms")
+        df_reports["timeOfNotification"] = pd.to_datetime(df_reports["timeOfNotification"], unit="ms")
+
+        print(f"ExtractAccelData initialized with time_window: {self.time_window}")
+        df_reports, df_accel = X
+        df_reports['accel_data'] = df_reports.apply(lambda row: self._extract_accel_data(row, df_accel), axis=1)
+    
+        print(f"Combining called with label_columns: {self.label_columns}")
+        combined_data = []
+
+
+        for _, row in df_reports.iterrows():
+            accel_data = row['accel_data']
+            for _, accel_row in accel_data.iterrows():
+                combined_row = {
+                    'participantId': row['participantId'],  # Participant ID
+                    'selfreport_time': row['timeOfNotification'],  # Self-report time
+                    'accel_time': accel_row['timeOfNotification'],  # Accelerometer data time
+                    'x': accel_row['x'],  # x-axis accelerometer data
+                    'y': accel_row['y'],  # y-axis accelerometer data
+                    'z': accel_row['z']   # z-axis accelerometer data
+                }
+
+                # Dynamically add emotion labels to the combined row
+                for label in self.label_columns:
+                    # print(f"Processing label: {label}")
+                    combined_row[label] = row[label]
+
+                combined_data.append(combined_row)
+
+        combined_df = pd.DataFrame(combined_data)
+
+        # Create groupid column (unique identifier based on participantId and selfreport_time)
+        combined_df['groupid'] = combined_df.groupby(['participantId', 'selfreport_time']).ngroup() + 1
+        col = combined_df.pop("groupid")  # Move groupid to the first column
+        combined_df.insert(0, col.name, col)
+
+        # Export the combined dataframe to CSV
+        time_window_str = str(self.time_window)
+        label_columns_str = "_".join(self.label_columns)
+        file_name = f"combined_data_timewindow_{time_window_str}min_labels_{label_columns_str}.csv"
+        combined_df.to_csv(file_name, index=False)
+        print(f"Combined dataframe exported successfully to {file_name}.")
+
+        return combined_df
+    
+    def _extract_accel_data(self, row, accel_data):
+        time_delta = pd.Timedelta(minutes=self.time_window)         
+        start_time = pd.to_datetime(row['timeOfNotification']) - time_delta
+        end_time = pd.to_datetime(row['timeOfNotification']) + time_delta
+        participant_id = row['participantId']
+        mask = (
+            (accel_data['participantId'] == participant_id) &       # Filter out rows with different participantId
+            (accel_data['timeOfNotification'] >= start_time) &      # Filter out rows with time outside the window
+            (accel_data['timeOfNotification'] <= end_time)          # Filter out rows with time outside the window
+        )
+        return accel_data[mask] 

pipeline_classes/extract_features.py

@@ -0,0 +1,272 @@
+import pandas as pd
+import numpy as np
+from sklearn.base import BaseEstimator, TransformerMixin
+from scipy.fftpack import fft
+from scipy.signal import welch
+import pywt
+from _config import config
+
+class ExtractFeatures(BaseEstimator, TransformerMixin):
+    def __init__(self, window_length, window_step_size, data_frequency, selected_domains=None, include_magnitude=False, label_columns=None):
+        self.window_length = window_length
+        self.window_step_size = window_step_size
+        self.data_frequency = data_frequency
+        self.selected_domains = selected_domains
+        self.include_magnitude = include_magnitude
+        self.label_columns = label_columns #if label_columns else ["arousal", "valence"]  # Default to arousal and valence if not specified
+
+    def fit(self, X, y=None):
+        return self
+
+    def transform(self, X):
+        features_list = []
+
+        if 'groupid' in X.columns:  # Check for groupid column
+            for groupid in X['groupid'].unique():  # Iterate over unique group IDs
+                temp = X[X['groupid'] == groupid]  # Filter rows by group ID
+                temp_ex = temp[['accel_time', 'x', 'y', 'z']].copy()  # Keep only the necessary columns (accel_time can be removed if unused)
+                windows = self._window_data(temp_ex[['x', 'y', 'z']])  # Create windows of data
+                
+                for window in windows:
+                    features = self._extract_features_from_window(window)  # Extract features from each window
+                    features['groupid'] = groupid  # Add groupid to the features
+                    
+                    # Dynamically add emotion labels to the features
+                    for label in self.label_columns:
+                        features[label] = temp[label].iloc[0]
+                    
+                    features_list.append(pd.DataFrame([features]))  # Convert dictionary to DataFrame
+                    
+        else:  # In case there's no groupid, calculate features without it
+            windows = self._window_data(X[['x', 'y', 'z']])
+            for window in windows:
+                features = self._extract_features_from_window(window)
+                features_list.append(pd.DataFrame([features]))
+
+        all_features = pd.concat(features_list, ignore_index=True)
+
+        # Export features to CSV
+        window_length_str = str(self.window_length)
+        window_step_size_str = str(self.window_step_size)
+        if self.selected_domains is None:  # All features calculated if domains are not selected
+            domain_str = "all_features"
+        else:
+            domain_str = "_".join(self.selected_domains)
+        file_name = f"features_window_{window_length_str}_step_{window_step_size_str}_{domain_str}.csv"
+        all_features.to_csv(file_name, index=False)
+
+        print("All features extracted successfully.")
+        return all_features
+
+    # Time Domain Features
+    def _calculate_magnitude(self, window):
+        return np.sqrt(window[:, 0]**2 + window[:, 1]**2 + window[:, 2]**2)
+    
+    def _window_data(self, data):                                                            # Function to create windows of the data
+        window_samples = int(self.window_length * self.data_frequency)                       # Number of samples in each window 60sec * 25Hz = 1500 samples
+        step_samples = int(self.window_step_size * self.data_frequency)                                             # Number of samples to move the window
+        windows = [data[i:i + window_samples] for i in range(0, len(data) - window_samples + 1, step_samples)]      # Create windows
+        return np.array(windows)
+
+    def _extract_features_from_window(self, window):                        #DONE Mehrere domains gleichzeitig berechnen
+        all_features = {}
+
+        if self.selected_domains is None or 'time_domain' in self.selected_domains:
+            all_features.update(self._extract_time_domain_features(window))
+        
+        if self.selected_domains is None or 'spatial' in self.selected_domains:
+            all_features.update(self._extract_spatial_features(window))
+        
+        if self.selected_domains is None or 'frequency' in self.selected_domains:
+            all_features.update(self._extract_frequency_domain_features(window))
+
+        if self.selected_domains is None or 'statistical' in self.selected_domains:
+            all_features.update(self._extract_statistical_features(window))
+
+        if self.selected_domains is None or 'wavelet' in self.selected_domains:
+            all_features.update(self._extract_wavelet_features(window))
+
+        return all_features
+
+    def _extract_time_domain_features(self, window):
+        features = {
+            'mean_x': np.mean(window[:, 0]),
+            'mean_y': np.mean(window[:, 1]),
+            'mean_z': np.mean(window[:, 2]),
+            'std_x': np.std(window[:, 0]),
+            'std_y': np.std(window[:, 1]),
+            'std_z': np.std(window[:, 2]),
+            'variance_x': np.var(window[:, 0]),
+            'variance_y': np.var(window[:, 1]),
+            'variance_z': np.var(window[:, 2]),
+            'rms_x': np.sqrt(np.mean(window[:, 0]**2)),
+            'rms_y': np.sqrt(np.mean(window[:, 1]**2)),
+            'rms_z': np.sqrt(np.mean(window[:, 2]**2)),
+            'max_x': np.max(window[:, 0]),
+            'max_y': np.max(window[:, 1]),
+            'max_z': np.max(window[:, 2]),
+            'min_x': np.min(window[:, 0]),
+            'min_y': np.min(window[:, 1]),
+            'min_z': np.min(window[:, 2]),
+            'peak_to_peak_x': np.ptp(window[:, 0]),
+            'peak_to_peak_y': np.ptp(window[:, 1]),
+            'peak_to_peak_z': np.ptp(window[:, 2]),
+            'skewness_x': pd.Series(window[:, 0]).skew(),
+            'skewness_y': pd.Series(window[:, 1]).skew(),
+            'skewness_z': pd.Series(window[:, 2]).skew(),
+            'kurtosis_x': pd.Series(window[:, 0]).kurt(),
+            'kurtosis_y': pd.Series(window[:, 1]).kurt(),
+            'kurtosis_z': pd.Series(window[:, 2]).kurt(),
+            'zero_crossing_rate_x': np.sum(np.diff(np.sign(window[:, 0])) != 0),
+            'zero_crossing_rate_y': np.sum(np.diff(np.sign(window[:, 1])) != 0),
+            'zero_crossing_rate_z': np.sum(np.diff(np.sign(window[:, 2])) != 0),
+            'sma' : np.sum(np.abs(window[:, 0])) + np.sum(np.abs(window[:, 1])) + np.sum(np.abs(window[:, 2])), #Signal Magnitude Area
+        }
+        # print(f"Time domain features extracted successfully.")
+
+        # Additional features for Magnitude (xyz in one vector)
+        if self.include_magnitude:
+            magnitude = self._calculate_magnitude(window)
+            features['mean_magnitude'] = np.mean(magnitude)
+            features['std_magnitude'] = np.std(magnitude)
+            features['variance_magnitude'] = np.var(magnitude)
+            features['rms_magnitude'] = np.sqrt(np.mean(magnitude**2))
+            features['max_magnitude'] = np.max(magnitude)
+            features['min_magnitude'] = np.min(magnitude)
+            features['peak_to_peak_magnitude'] = np.ptp(magnitude)
+            features['skewness_magnitude'] = pd.Series(magnitude).skew()
+            features['kurtosis_magnitude'] = pd.Series(magnitude).kurt()
+            features['zero_crossing_rate_magnitude'] = np.sum(np.diff(np.sign(magnitude)) != 0)
+            # print(f"Additional time domain features for magnitude extracted successfully.")
+
+        return features
+
+    # Spatial Features
+    def _extract_spatial_features(self, window):
+        features = {}
+
+        # Euclidean Norm (Magnitude)
+        magnitude = self._calculate_magnitude(window)
+        features['euclidean_norm'] = np.mean(magnitude)  # or np.linalg.norm for each window
+
+        # Tilt Angles (Pitch and Roll)
+        pitch = np.arctan2(window[:, 1], np.sqrt(window[:, 0]**2 + window[:, 2]**2)) * (180 / np.pi)
+        roll = np.arctan2(window[:, 0], np.sqrt(window[:, 1]**2 + window[:, 2]**2)) * (180 / np.pi)
+        features['mean_pitch'] = np.mean(pitch)
+        features['mean_roll'] = np.mean(roll)
+
+        # Correlation between Axes
+        features['correlation_xy'] = np.corrcoef(window[:, 0], window[:, 1])[0, 1]
+        features['correlation_xz'] = np.corrcoef(window[:, 0], window[:, 2])[0, 1]
+        features['correlation_yz'] = np.corrcoef(window[:, 1], window[:, 2])[0, 1]
+
+        # print(f"Spatial features extracted successfully.")
+        return features
+
+
+
+    # Frequency Domain Features
+    def _extract_frequency_domain_features(self, window):
+        n = len(window)
+        freq_values = np.fft.fftfreq(n, d=1/self.data_frequency)[:n // 2]
+        fft_values = fft(window, axis=0)
+        fft_magnitude = np.abs(fft_values)[:n // 2]
+
+        features = {}
+
+        # Spectral Entropy
+        def spectral_entropy(signal):
+            psd = np.square(signal)
+            psd_norm = psd / np.sum(psd)
+            return -np.sum(psd_norm * np.log(psd_norm + 1e-10))
+
+        for i, axis in enumerate(['x', 'y', 'z']):
+            # Dominant Frequency
+            dominant_frequency = freq_values[np.argmax(fft_magnitude[:, i])]
+            features[f'dominant_frequency_{axis}'] = dominant_frequency
+
+            # Spectral Entropy
+            entropy = spectral_entropy(fft_magnitude[:, i])
+            features[f'spectral_entropy_{axis}'] = entropy
+
+            # Power Spectral Density (PSD) and Energy
+            f, psd_values = welch(window[:, i], fs=self.data_frequency, nperseg=n)
+            features[f'psd_mean_{axis}'] = np.mean(psd_values)
+            features[f'energy_{axis}'] = np.sum(psd_values**2)
+
+            # Bandwidth (frequency range containing significant portion of the energy)
+            cumulative_energy = np.cumsum(psd_values)
+            total_energy = cumulative_energy[-1]
+            low_cutoff_idx = np.argmax(cumulative_energy > 0.1 * total_energy)
+            high_cutoff_idx = np.argmax(cumulative_energy > 0.9 * total_energy)
+            bandwidth = f[high_cutoff_idx] - f[low_cutoff_idx]
+            features[f'bandwidth_{axis}'] = bandwidth
+
+            # Spectral Centroid (Center of mass of the spectrum)
+            spectral_centroid = np.sum(f * psd_values) / np.sum(psd_values)
+            features[f'spectral_centroid_{axis}'] = spectral_centroid
+
+        if self.include_magnitude:
+            # Magnitude-based Frequency Domain Features
+            magnitude = self._calculate_magnitude(window)
+            fft_magnitude_mag = np.abs(fft(magnitude))[:n // 2]
+
+            # Dominant Frequency for Magnitude
+            features['dominant_frequency_magnitude'] = freq_values[np.argmax(fft_magnitude_mag)]
+
+            # Spectral Entropy for Magnitude
+            features['spectral_entropy_magnitude'] = spectral_entropy(fft_magnitude_mag)
+
+            # Power Spectral Density and Energy for Magnitude
+            f, psd_values_mag = welch(magnitude, fs=self.data_frequency, nperseg=n)
+            features['psd_mean_magnitude'] = np.mean(psd_values_mag)
+            features['energy_magnitude'] = np.sum(psd_values_mag**2)
+
+            # Bandwidth for Magnitude
+            cumulative_energy_mag = np.cumsum(psd_values_mag)
+            total_energy_mag = cumulative_energy_mag[-1]
+            low_cutoff_idx_mag = np.argmax(cumulative_energy_mag > 0.1 * total_energy_mag)
+            high_cutoff_idx_mag = np.argmax(cumulative_energy_mag > 0.9 * total_energy_mag)
+            bandwidth_mag = f[high_cutoff_idx_mag] - f[low_cutoff_idx_mag]
+            features['bandwidth_magnitude'] = bandwidth_mag
+
+            # Spectral Centroid for Magnitude
+            features['spectral_centroid_magnitude'] = np.sum(f * psd_values_mag) / np.sum(psd_values_mag)
+
+        # print(f"Frequency domain features extracted successfully.")
+        return features
+
+
+    def _extract_statistical_features(self, window):
+        features = {
+            '25th_percentile_x': np.percentile(window[:, 0], 25),
+            '25th_percentile_y': np.percentile(window[:, 1], 25),
+            '25th_percentile_z': np.percentile(window[:, 2], 25),
+            '75th_percentile_x': np.percentile(window[:, 0], 75),
+            '75th_percentile_y': np.percentile(window[:, 1], 75),
+            '75th_percentile_z': np.percentile(window[:, 2], 75),
+        }
+        
+        if self.include_magnitude:
+            magnitude = self._calculate_magnitude(window)
+            features['25th_percentile_magnitude'] = np.percentile(magnitude, 25)
+            features['75th_percentile_magnitude'] = np.percentile(magnitude, 75)
+        
+        # print(f"Statistical features extracted successfully.")
+        return features
+
+    def _extract_wavelet_features(self, window, wavelet='db1'):
+        coeffs = pywt.wavedec(window, wavelet, axis=0, level=3)
+        features = {
+            'wavelet_energy_approx_x': np.sum(coeffs[0][:, 0]**2),
+            'wavelet_energy_approx_y': np.sum(coeffs[0][:, 1]**2),
+            'wavelet_energy_approx_z': np.sum(coeffs[0][:, 2]**2),
+        }
+        
+        if self.include_magnitude:
+            magnitude = self._calculate_magnitude(window)
+            coeffs_magnitude = pywt.wavedec(magnitude, wavelet, level=3)
+            features['wavelet_energy_approx_magnitude'] = np.sum(coeffs_magnitude[0]**2)
+        
+        # print(f"Wavelet features extracted successfully.")
+        return features

pipeline_classes/import_data.py

@@ -0,0 +1,51 @@
+import pandas as pd
+import numpy as np
+from sklearn.base import BaseEstimator, TransformerMixin
+from _config import config
+
+class ImportData(BaseEstimator, TransformerMixin):
+    def __init__(self, use_accel=True, use_reports=True, use_combined=False, use_features=False):
+        self.use_accel = use_accel
+        self.use_reports = use_reports
+        self.use_combined = use_combined
+        self.use_features = use_features
+        # Dynamically assign file paths based on usage
+        self.accel_path = config["accel_path"] if use_accel else None
+        self.reports_path = config["reports_path"] if use_reports else None
+        self.combined_data_path = config["combined_data_path"] if use_combined else None
+        self.features_data_path = config["features_data_path"] if use_features else None
+
+    def fit(self, X, y=None):
+        return self
+
+    def transform(self, X):
+        if self.features_data_path:  # If the path to features data is provided
+            # Load the features dataframe
+            features_df = pd.read_csv(self.features_data_path)
+            print('Features dataframe imported successfully.')
+            return features_df
+
+        elif self.combined_data_path:  # If the path to combined data is provided
+            # Load the combined dataframe
+            combined_df = pd.read_csv(self.combined_data_path)
+            print('Combined dataframe imported successfully.')
+            return combined_df
+
+        else:  # Otherwise, load the raw accelerometer and reports data
+            if not self.accel_path:
+                raise ValueError("accel_path needs to be provided if combined_data_path and features_data_path are not given.")
+            
+            # Load accelerometer data
+            raw_acceleration_data = pd.read_csv(self.accel_path)
+            df_accel = raw_acceleration_data.copy()
+
+            if self.reports_path:
+                # Load self-reports data if provided
+                raw_selfreports_data = pd.read_csv(self.reports_path)
+                df_reports = raw_selfreports_data.copy()
+                print('Raw data (accelerometer and self-reports) imported successfully.')
+                return df_reports, df_accel
+            else:
+                # Only accelerometer data provided
+                print('Raw accelerometer data imported successfully.')
+                return df_accel

pipeline_classes/lowpassfilter.py

@@ -0,0 +1,55 @@
+import numpy as np
+import pandas as pd
+from sklearn.base import BaseEstimator, TransformerMixin
+from scipy.signal import butter, filtfilt
+
+class LowPassFilter(BaseEstimator, TransformerMixin):
+    def __init__(self, cutoff_frequency, sampling_rate, order):
+        """
+        Initialize the LowPassFilter class.
+        
+        Parameters:
+        - cutoff_frequency: The cutoff frequency for the low-pass filter (default: 5 Hz).
+        - sampling_rate: The sampling rate of the accelerometer data (default: 25 Hz).
+        - order: The order of the filter (default: 4).
+        """
+        self.cutoff_frequency = cutoff_frequency
+        self.sampling_rate = sampling_rate
+        self.order = order
+
+    def _butter_lowpass_filter(self, data):
+        """
+        Apply a Butterworth low-pass filter to the data.
+        
+        Parameters:
+        - data: A NumPy array containing the accelerometer data to be filtered.
+        
+        Returns:
+        - A filtered NumPy array.
+        """
+        nyquist = 0.5 * self.sampling_rate
+        normalized_cutoff = self.cutoff_frequency / nyquist
+        b, a = butter(self.order, normalized_cutoff, btype='low', analog=False)
+        filtered_data = filtfilt(b, a, data, axis=0)
+        return filtered_data
+
+    def fit(self, X, y=None):
+        return self
+
+    def transform(self, X):
+        """
+        Apply the low-pass filter to the accelerometer data.
+        
+        Parameters:
+        - X: A DataFrame with 'x', 'y', and 'z' columns representing the accelerometer data.
+        
+        Returns:
+        - The DataFrame with filtered 'x', 'y', and 'z' columns.
+        """
+        if 'x' in X.columns and 'y' in X.columns and 'z' in X.columns:
+            X[['x', 'y', 'z']] = self._butter_lowpass_filter(X[['x', 'y', 'z']].values)
+            print("Low-pass filter applied successfully.")
+        else:
+            raise ValueError("The input DataFrame must contain 'x', 'y', and 'z' columns.")
+        
+        return X

pipeline_classes/pcahandler.py

@@ -0,0 +1,24 @@
+import pandas as pd
+import numpy as np
+from sklearn.base import BaseEstimator, TransformerMixin
+from sklearn.decomposition import PCA
+from _config import config
+    
+class PCAHandler(BaseEstimator, TransformerMixin):
+    def __init__(self, apply_pca=False, variance=0.95):
+        self.apply_pca = apply_pca
+        self.variance = variance
+        self.pca = None
+
+    def fit(self, X, y=None):
+        if self.apply_pca:
+            self.pca = PCA(n_components=self.variance)
+            self.pca.fit(X)
+        return self
+
+    def transform(self, X):
+        if self.apply_pca and self.pca:
+            X_transformed = self.pca.transform(X)
+            return pd.DataFrame(X_transformed, index=X.index)
+        
+        return X

pipeline_classes/scale_xyzdata.py

@@ -0,0 +1,28 @@
+import pandas as pd
+import numpy as np
+from sklearn.base import BaseEstimator, TransformerMixin
+from sklearn.preprocessing import StandardScaler, MinMaxScaler
+from _config import config
+
+class ScaleXYZData(BaseEstimator, TransformerMixin):
+    def __init__(self, scaler_type='standard'):
+        self.scaler_type = scaler_type
+
+    def fit(self, X, y=None):
+        return self
+
+    def transform(self, X):
+        columns_to_scale = ['x', 'y', 'z']                  
+        if self.scaler_type == 'standard':                  # Scale the columns using StandardScaler or MinMaxScaler
+            scaler = StandardScaler()
+        elif self.scaler_type == 'minmax':
+            scaler = MinMaxScaler()
+        elif self.scaler_type == 'none':
+            return X  # Return the DataFrame without scaling
+        else:
+            raise ValueError("Invalid scaler_type. Expected 'standard' or 'minmax'.")   # Raise an error if scaler_type is invalid
+        scaled_columns = scaler.fit_transform(X[columns_to_scale])
+        scaled_df = pd.DataFrame(scaled_columns, columns=columns_to_scale, index=X.index)
+        X[columns_to_scale] = scaled_df
+        print("Data scaled successfully.")
+        return X

pipeline_classes/train_model.py

@@ -0,0 +1,179 @@
+import pandas as pd
+import numpy as np
+from sklearn.base import BaseEstimator, TransformerMixin
+from sklearn.model_selection import StratifiedGroupKFold
+from skopt import BayesSearchCV
+from sklearn.ensemble import RandomForestClassifier
+from sklearn.svm import SVC
+from xgboost import XGBClassifier
+import joblib
+from skopt.space import Real, Integer, Categorical
+from sklearn.metrics import classification_report, accuracy_score
+import json
+from sklearn.preprocessing import LabelEncoder
+from _config import config   
+
+class TrainModel(BaseEstimator, TransformerMixin):
+    def __init__(self, config):
+        self.config = config
+        self.target = config.get("target_label", None)  # User-defined target label in config
+        self.label_encoder = LabelEncoder()
+        self.selected_domains = self.config.get("selected_domains", "All domains")  # Default to all domains if None
+
+        if not self.target:
+            raise ValueError("No target label specified in the config. Please set 'target_label'.")
+
+    def get_default_param_space(self, classifier):
+        """ Returns the default hyperparameter space for a given classifier. """
+        if classifier == 'xgboost':
+            return {
+                'learning_rate': Real(0.01, 0.3, prior='log-uniform'),
+                'n_estimators': Integer(100, 1000),
+                'max_depth': Integer(3, 10),
+                'min_child_weight': (1, 10),
+                'subsample': (0.5, 1.0),
+                'colsample_bytree': (0.5, 1.0),
+                'gamma': (0, 10),
+                'reg_alpha': (0, 10),
+                'reg_lambda': (0, 10),
+            }
+        elif classifier == 'svm':
+            return {
+                'C': Real(0.1, 10, prior='log-uniform'),
+                'kernel': Categorical(['linear', 'rbf'])
+            }
+        elif classifier == 'randomforest':
+            return {
+                'n_estimators': Integer(100, 1000),
+                'max_depth': Integer(3, 10)
+            }
+        else:
+            raise ValueError(f"Unsupported classifier type: {classifier}")
+
+    def fit(self, X, y=None):
+        # Ensure the target column exists in the dataset
+        if self.target not in X.columns:
+            raise ValueError(f"Target label '{self.target}' not found in the dataset.")
+        
+        # Fit the label encoder on the target column
+        print(f"Encoding the target labels for '{self.target}'...")
+        self.label_encoder.fit(X[self.target])
+
+        # Print the mapping between original labels and encoded labels
+        original_labels = list(self.label_encoder.classes_)
+        encoded_labels = list(range(len(original_labels)))
+        label_mapping = dict(zip(encoded_labels, original_labels))
+        print(f"Label encoding complete. Mapping: {label_mapping}")
+
+        # Transform the target column and add it as 'encoded_target'
+        X['encoded_target'] = self.label_encoder.transform(X[self.target])
+
+        # Value counts for the encoded target        
+        value_counts = X['encoded_target'].value_counts().to_dict()
+        print(f"Value counts for encoded target: {value_counts}")
+        
+        # Pop unnecessary columns (groupid, emotion labels not being used, etc.)
+        groups = X.pop('groupid')
+        # Pop the label columns which aren't used
+        for label in self.config["label_columns"]:
+                X.pop(label)
+
+        # Pop the encoded target as Y
+        y = X.pop('encoded_target')
+
+        # Store the feature names for later use
+        feature_names = X.columns.tolist()
+
+        # Choose classifier
+        classifier = self.config['classifier']
+        if classifier == 'xgboost':
+            model = XGBClassifier(objective='multi:softmax', random_state=42)
+        elif classifier == 'svm':
+            model = SVC(probability=True)
+        elif classifier == 'randomforest':
+            model = RandomForestClassifier(random_state=42)
+        else:
+            raise ValueError(f"Unsupported classifier type: {classifier}")
+
+        print(f"Training the model using {classifier}...")
+
+        # Use user-defined param_space if provided, otherwise use default
+        print(f"Classifier: {classifier}")
+        default_param_space = self.get_default_param_space(classifier)
+        param_space = self.config.get("param_space") or default_param_space     
+
+        # Hyperparameter tuning using Bayesian optimization
+        sgkf = StratifiedGroupKFold(n_splits=self.config['n_splits'])
+        print(f"Parameter space being used: {param_space}")
+        if param_space is None:
+            raise ValueError("Parameter space cannot be None. Please check the classifier configuration.")
+
+        opt = BayesSearchCV(
+            estimator=model,
+            search_spaces=param_space,
+            cv=sgkf,
+            n_iter=self.config['n_iter'],
+            n_jobs=self.config['n_jobs'],
+            n_points=self.config['n_points'],
+            verbose=1,
+            scoring='accuracy'
+        )
+
+        print("Hyperparameter tuning in progress...")
+        print(X.describe(),X.columns)
+        print(f"stop")
+
+        # Fit the model using the encoded target
+        opt.fit(X, y, groups=groups)
+        self.best_model = opt.best_estimator_
+        print(f"Best parameters found: {opt.best_params_}")
+
+        # Print classification metrics
+        y_pred = self.best_model.predict(X)
+        accuracy = accuracy_score(y, y_pred)
+        report = classification_report(y, y_pred, target_names=self.label_encoder.classes_, output_dict=True)
+        
+        # Save classification report
+        classification_report_json = report  
+        with open(f'classification_report_{self.target}.json', 'w') as f:
+            json.dump(classification_report_json, f, indent=4)
+
+        print(f"Accuracy: {accuracy}")
+        print(f"Classification Report:\n{report}")
+
+        # Save the best model with the target label in the file name
+        model_name = f"{classifier}_best_model_{self.target}.pkl"
+        joblib.dump(self.best_model, model_name)
+        print("Model saved successfully.")
+
+        # Save model metadata
+        model_metadata = {
+            "best_params": opt.best_params_,
+            "accuracy": accuracy,
+            "classification_report": classification_report_json,
+            "label_mapping": label_mapping,
+            "model_name": model_name,
+            "value_counts": value_counts,
+            "selected_domains": self.selected_domains,  
+            "include_magnitude": self.config.get("include_magnitude", True)      
+        }
+
+        if hasattr(self.best_model, "feature_importances_"):
+            feature_importances = self.best_model.feature_importances_
+            # Convert feature importances to native Python floats
+            feature_importance_dict = {feature: float(importance) for feature, importance in zip(feature_names, feature_importances)}
+            model_metadata["feature_importances"] = feature_importance_dict
+            print("Feature Importances:")
+            for feature, importance in feature_importance_dict.items():
+                print(f"{feature}: {importance:.4f}")
+
+        # Save metadata with the target name in the file name
+        metadata_file = f"{classifier}_model_metadata_{self.target}.json"
+        with open(metadata_file, "w") as f:
+            json.dump(model_metadata, f, indent=4)
+            print(f"Model metadata saved to {metadata_file}.")
+
+        return self
+
+    def transform(self, X):
+        return X  # Placeholder for transform step (not needed for training)

requirements.txt

@@ -0,0 +1,8 @@
+pandas>=1.3.0
+numpy>=1.21.0
+scikit-learn>=1.0.0
+scikit-optimize>=0.9.0
+xgboost>=1.5.0
+joblib>=1.0.0
+PyWavelets>=1.2.0
+scipy>=1.7.0