File size: 19,025 Bytes
ca3e099
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
import gradio as gr
from pipeline_classes import CreateCombinedDataFrame, ScaleXYZData, ExtractFeatures, TrainModel, ClassifyMovementData, LowPassFilter, PCAHandler
from sklearn.pipeline import Pipeline
#from _config import config
import pandas as pd
import numpy as np
import joblib
import json


# Define pipelines
combining_dataframes_pipeline = Pipeline([
    #('import_data', ImportData(use_accel=True, use_reports=True, use_combined=False, use_features=False)),
    ('create_combined_dataframe', CreateCombinedDataFrame(time_window=None, label_columns=None)),
])

feature_extraction_pipeline = Pipeline([
    #('import_data', ImportData(use_accel=False, use_reports=False, use_combined=True, use_features=False)),
    ('low_pass_filter', LowPassFilter(cutoff_frequency=None, sampling_rate=None, order=None)),
    ('scale_xyz_data', ScaleXYZData(scaler_type=None)),
    ('extract_features', ExtractFeatures(window_length=None,
                                         window_step_size=None,
                                         data_frequency=None,
                                         selected_domains=None,
                                         include_magnitude=None,
                                         features_label_columns=None)),
])

training_model_pipeline = Pipeline([
    #('import_data', ImportData(use_accel=False, use_reports=False, use_combined=False, use_features=True)),
    ('pca_handler', PCAHandler(apply_pca=None, variance=None)),
    ('train_model', TrainModel(classifier=None, train_label= None, target=None)),
])

analyzing_data_pipeline = Pipeline([
    #('import_data', ImportData(use_accel=True, use_reports=False, use_combined=False, use_features=False)),
    ('low_pass_filter', LowPassFilter(cutoff_frequency=None, sampling_rate=None, order=None)),
    ('scale_xyz_data', ScaleXYZData(scaler_type=None)),
    ('extract_features', ExtractFeatures(window_length=None,
                                         window_step_size=None,
                                         data_frequency=None,
                                         selected_domains=None,
                                         include_magnitude=None,
                                         features_label_columns=None)),
    ('classify_movement_data', ClassifyMovementData(model_file=None)),
])

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=None, label_columns=None)),
    ('low_pass_filter', LowPassFilter(cutoff_frequency=None, sampling_rate=None, order=None)),
    ('scale_xyz_data', ScaleXYZData(scaler_type=None)),
    ('extract_features', ExtractFeatures(window_length=None,
                                         window_step_size=None,
                                         data_frequency=None,
                                         selected_domains=None,
                                         include_magnitude=None,
                                         features_label_columns=None)),
    ('pca_handler', PCAHandler(apply_pca=None, variance=None)),
    ('train_model', TrainModel(classifier=None, train_label= None, target=None)),
])

def execute_combine_pipeline(accel_file, report_file,
                     time_window=None, label_columns=None
                     ):
    try:      
        # Load data files only if paths are valid
        accel_data = pd.read_csv(accel_file) if accel_file else None
        report_data = pd.read_csv(report_file) if report_file else None

        # Validate inputs for the selected pipeline
        if accel_data is None or report_data is None:
            return "Error: Both accelerometer and self-report data files are required for this pipeline.", None
        combining_dataframes_pipeline.set_params(
            create_combined_dataframe__time_window=time_window,
            create_combined_dataframe__label_columns=label_columns.split(','))
        X = report_data, accel_data
        result = combining_dataframes_pipeline.fit_transform(X)
        output_file = "combine_dataframes_output.csv"
        result.to_csv(output_file, index=False)

        return output_file
    
    except Exception as e:
        print(f"Error occurred: {str(e)}")
        return str(e), None


def execute_feature_extraction_pipeline(combined_file, cutoff_frequency, order, scaler_type, window_length, window_step_size, data_frequency, include_magnitude, features_label_columns):
    try:
        combined_data = pd.read_csv(combined_file) if combined_file else None
        if combined_data is None:
            return "Error: Combined data file is required for this pipeline.", None

        feature_extraction_pipeline.set_params(
            low_pass_filter__cutoff_frequency=cutoff_frequency,
            low_pass_filter__order=order, 
            low_pass_filter__sampling_rate=data_frequency,
            scale_xyz_data__scaler_type=scaler_type, 
            extract_features__window_length=window_length,
            extract_features__window_step_size=window_step_size, 
            extract_features__data_frequency=data_frequency,
            #extract_features__selected_domains=None,
            extract_features__include_magnitude=include_magnitude,
            extract_features__features_label_columns=features_label_columns.split(','))
        result = feature_extraction_pipeline.fit_transform(combined_data)
        output_file = "extract_features_output.csv"
        result.to_csv(output_file, index=False)
        return output_file

    except Exception as e:
        print(f"Error occurred: {str(e)}")
        return str(e)
    
def execute_training_pipeline(features_file, apply_pca, pca_variance, classifier, train_label, target):
    try:
        print(f"features_file: {features_file}")
        features_data = pd.read_csv(features_file) if features_file else None
        if features_data is None:
            return "Error: Features data file is required for this pipeline.", None
        
        training_model_pipeline.set_params(
            pca_handler__apply_pca=apply_pca,
            pca_handler__variance=pca_variance,
            train_model__classifier=classifier,
            train_model__train_label=train_label,
            train_model__target=target)
        
        X = features_data
        training_model_pipeline.fit(X)
        output_file, secondary_output_file = training_model_pipeline.named_steps['train_model'].get_output_files()
        return output_file, secondary_output_file

    except Exception as e:
        print(f"Error occurred: {str(e)}")
        return str(e), None
    
def execute_analyze_pipeline(accel_file, model_file, cutoff_frequency, order, scaler_type, window_length, data_frequency, include_magnitude, features_label_columns):
    try:
        accel_data = pd.read_csv(accel_file) if accel_file else None
        if accel_data is None:
            return "Error: Accelerometer data file is required for this pipeline.", None
        
        analyzing_data_pipeline.set_params(
            low_pass_filter__cutoff_frequency=cutoff_frequency,
            low_pass_filter__order=order, 
            low_pass_filter__sampling_rate=data_frequency,
            scale_xyz_data__scaler_type=scaler_type, 
            extract_features__window_length=window_length,
            extract_features__window_step_size=window_length, 
            extract_features__data_frequency=data_frequency,
            #extract_features__selected_domains=None,
            extract_features__include_magnitude=include_magnitude,
            extract_features__features_label_columns=features_label_columns.split(','),
            classify_movement_data__model_file=model_file.name
            )

        result = analyzing_data_pipeline.fit_transform(accel_data)
        output_file = "analyze_data_output.csv"
        result.to_csv(output_file, index=False)
        return output_file

    except Exception as e:
        print(f"Error occurred: {str(e)}")
        return str(e), None
    
def execute_complete_training_pipeline(accel_file, report_file, time_window, label_columns,
                                       cutoff_frequency, order, scaler_type, window_length, window_step_size, data_frequency, include_magnitude, features_label_columns,
                                        apply_pca, pca_variance, classifier, train_label, target):
    try:
        accel_data = pd.read_csv(accel_file) if accel_file else None
        report_data = pd.read_csv(report_file) if report_file else None
        if accel_data is None or report_data is None:
            return "Error: Both accelerometer and self-report data files are required for this pipeline.", None
        
        complete_training_model_pipeline.set_params(
            create_combined_dataframe__time_window=time_window,
            create_combined_dataframe__label_columns=label_columns.split(','),
            low_pass_filter__cutoff_frequency=cutoff_frequency,
            low_pass_filter__order=order, 
            low_pass_filter__sampling_rate=data_frequency,
            scale_xyz_data__scaler_type=scaler_type, 
            extract_features__window_length=window_length,
            extract_features__window_step_size=window_step_size, 
            extract_features__data_frequency=data_frequency,
            #extract_features__selected_domains=None,
            extract_features__include_magnitude=include_magnitude,
            extract_features__features_label_columns=label_columns.split(','),
            pca_handler__apply_pca=apply_pca,
            pca_handler__variance=pca_variance,
            train_model__classifier=classifier,
            train_model__train_label=label_columns,
            train_model__target=target
        )
        X = report_data, accel_data
        complete_training_model_pipeline.fit(X)
        output_file, secondary_output_file = complete_training_model_pipeline.named_steps['train_model'].get_output_files()
        return output_file, secondary_output_file

    except Exception as e:
        print(f"Error occurred: {str(e)}")
        return str(e), None

# Gradio Blocks Interface
with gr.Blocks() as demo:
    with gr.Tabs():
        with gr.TabItem("Combine DataFrames"):
            accel_file = gr.File(label="Upload Accelerometer Data")
            report_file = gr.File(label="Upload Self-Report Data")
            time_window = gr.Number(label="Time Window (minutes)", value=2)
            label_columns = gr.Textbox(label="Label Columns (comma-separated)", value="valence,arousal")
            combine_button = gr.Button("Combine DataFrames")
            combine_output = gr.File(label="Download Combined DataFrame")

            def combine_dataframes(accel_file, report_file, time_window, label_columns):
                output_file = execute_combine_pipeline(accel_file, report_file, time_window, label_columns)
                return output_file

            combine_button.click(combine_dataframes, inputs=[accel_file, report_file, time_window, label_columns], outputs=combine_output)

        with gr.TabItem("Extract Features"):
            combined_file = gr.File(label="Upload Combined Data")

            cutoff_frequency = gr.Number(label="Cutoff Frequency (Hz)", value=10)
            order = gr.Number(label="Order", value=4)

            scaler_type = gr.Radio(label="Scaler Type", choices=["standard", "minmax"])

            window_length = gr.Number(label="Window Length (seconds)", value=60)
            window_step_size = gr.Number(label="Window Step Size (seconds)", value=20)
            data_frequency = gr.Number(label="Data Frequency (Hz)", value=25)

            #selected_domains= gr.Textbox(label="Only these domains (Comma-Seperated) / If you want all then leave out", value=None)
            include_magnitude= gr.Checkbox(label="Include Magnitude", value=True)
            features_label_columns= gr.Textbox(label="Label Columns (comma-separated)", value="valence,arousal")
        
            extract_button = gr.Button("Extract Features")
            extract_output = gr.File(label="Download Extracted Features")

            def extract_features(combined_file, cutoff_frequency, order, scaler_type, window_length, window_step_size, data_frequency, include_magnitude, features_label_columns):
                    output_file = execute_feature_extraction_pipeline(combined_file,
                                                    cutoff_frequency, order, scaler_type, window_length, window_step_size, data_frequency,
                                                     include_magnitude, features_label_columns
                                                    )
                    return output_file

            extract_button.click(extract_features, inputs=[combined_file, cutoff_frequency, order, scaler_type, window_length, window_step_size,
                                                           data_frequency, include_magnitude, features_label_columns],  outputs=extract_output)

        with gr.TabItem("Train Model"):
            features_file = gr.File(label="Upload Features Data")

            apply_pca = gr.Checkbox(label="Apply PCA", value=False)
            pca_variance = gr.Number(label="PCA Variance", value=0.95)
            classifier = gr.Dropdown(label="Classifier", choices=["xgboost", "svm", "randomforest"], value="xgboost")
            train_label = gr.Textbox(label="Label Columns (comma-separated)", value="valence,arousal")
            target = gr.Textbox(label="Target Label", value="arousal")

            train_button = gr.Button("Train Model")
            train_output_json = gr.File(label="Download Model PKL")
            train_output_pkl = gr.File(label="Download Model JSON")

            def train_model(features_file, apply_pca, pca_variance, classifier, train_label,  target):
                output_file, secondary_output_file = execute_training_pipeline(features_file, apply_pca, pca_variance, classifier, train_label, target)
                return output_file, secondary_output_file

            train_button.click(train_model, inputs=[features_file, apply_pca, pca_variance, classifier, train_label, target], outputs=[train_output_json, train_output_pkl])

        with gr.TabItem("Analyze Data"):
            accel_file = gr.File(label="Upload Accelerometer Data")
            model_file = gr.File(label="Upload Model")

            cutoff_frequency = gr.Number(label="Cutoff Frequency (Hz)", value=10)
            order = gr.Number(label="Order", value=4)

            scaler_type = gr.Radio(label="Scaler Type", choices=["standard", "minmax"])

            window_length = gr.Number(label="Window Length (seconds)", value=60)
            data_frequency = gr.Number(label="Data Frequency (Hz)", value=25)

            #selected_domains= gr.Textbox(label="Only these domains (Comma-Seperated) / If you want all then leave out", value=None)
            include_magnitude= gr.Checkbox(label="Include Magnitude", value=True)
            features_label_columns= gr.Textbox(label="Label Columns (comma-separated)", value="valence,arousal")
        
            analyze_button = gr.Button("Analyze Data")
            analyze_output = gr.File(label="Download Analyzed Data")

            def analyze_data(accel_file, model_file, cutoff_frequency, order, scaler_type, window_length, data_frequency, include_magnitude, features_label_columns):
                output_file = execute_analyze_pipeline(accel_file, model_file, cutoff_frequency, order, scaler_type, window_length,
                                                           data_frequency, include_magnitude, features_label_columns)
                return output_file

            analyze_button.click(analyze_data, inputs=[accel_file, model_file, cutoff_frequency, order, scaler_type, window_length,
                                                           data_frequency, include_magnitude, features_label_columns ], outputs=analyze_output)

        with gr.TabItem("Complete Train Model"):
            accel_file = gr.File(label="Upload Accelerometer Data")
            report_file = gr.File(label="Upload Self-Report Data")

            time_window = gr.Number(label="Time Window (minutes)", value=2)
            label_columns = gr.Textbox(label="Label Columns (comma-separated)", value="valence,arousal")

            cutoff_frequency = gr.Number(label="Cutoff Frequency (Hz)", value=10)
            order = gr.Number(label="Order", value=4)

            scaler_type = gr.Radio(label="Scaler Type", choices=["standard", "minmax"])
            
            window_length = gr.Number(label="Window Length (seconds)", value=60)
            window_step_size = gr.Number(label="Window Step Size (seconds)", value=20)
            data_frequency = gr.Number(label="Data Frequency (Hz)", value=25)

            include_magnitude= gr.Checkbox(label="Include Magnitude", value=True)
            #features_label_columns= gr.Textbox(label="Label Columns (comma-separated)", value="valence,arousal")

            apply_pca = gr.Checkbox(label="Apply PCA", value=False)
            pca_variance = gr.Number(label="PCA Variance", value=0.95)
            classifier = gr.Dropdown(label="Classifier", choices=["xgboost", "svm", "randomforest"], value="xgboost")
            #train_label = gr.Textbox(label="Label Columns (comma-separated)", value="valence,arousal")
            target = gr.Textbox(label="Target Label", value="arousal")

            complete_train_button = gr.Button("Complete Train Model")

            complete_train_output_pkl = gr.File(label="Download Model PKL")
            complete_train_output_json = gr.File(label="Download Model JSON")

            def complete_train_model(accel_file, report_file, time_window, label_columns, 
                                       cutoff_frequency, order, scaler_type, window_length, window_step_size, data_frequency, include_magnitude, features_label_columns,
                                        apply_pca, pca_variance, classifier, train_label, target):
                output_file, secondary_output_file = execute_complete_training_pipeline(accel_file, report_file, time_window, label_columns, 
                                       cutoff_frequency, order, scaler_type, window_length, window_step_size, data_frequency, include_magnitude, features_label_columns,
                                        apply_pca, pca_variance, classifier, train_label, target)
                return output_file, secondary_output_file

            complete_train_button.click(complete_train_model, inputs=[accel_file, report_file, time_window, label_columns, 
                                       cutoff_frequency, order, scaler_type, window_length, window_step_size, data_frequency, include_magnitude, features_label_columns,
                                        apply_pca, pca_variance, classifier, train_label, target], outputs=[complete_train_output_pkl, complete_train_output_json])


demo.launch()