olfactory-pca-decoding/preprocessing.py at main · vivianzhou1288/olfactory-pca-decoding · GitHub

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
# preprocessing.py
from simulation import create_feature_matrix
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
import pandas as pd

def prepare_data(n_mice=5, n_trials_per_mouse=40, test_size=0.2, random_state=42):
    """
    Prepare data for ML models.

    Returns:
        X_train_scaled: Scaled training features
        X_test_scaled: Scaled test features
        y_train: Training labels (1=Correct, 0=Incorrect)
        y_test: Test labels
        feature_cols: List of feature column names
        scaler: Fitted StandardScaler object
    """

    # Generate data
    feature_matrix, trial_labels = create_feature_matrix(n_mice, n_trials_per_mouse, random_state=random_state)

    # Create target
    feature_matrix['Correct'] = (feature_matrix['Hit'] + feature_matrix['CR']).astype(int)

    # One-hot encode learning stage
    # Converts: "Naive", "Learning", "Proficient" → binary columns
    # drop_first=True: Uses Naive as reference (encoded as [0,0])
    feature_matrix_encoded = pd.get_dummies(feature_matrix,
                                            columns=['Learning_Stage'],
                                            drop_first=True)

    # Select features
    feature_cols = ['theta_frequency', 'beta_frequency', 'gamma_frequency',
                    'noise_level', 'Trial_Number',
                    'Learning_Stage_Naive', 'Learning_Stage_Proficient']

    X = feature_matrix_encoded[feature_cols]
    y = feature_matrix_encoded['Correct']

    # Train-test split
    X_train, X_test, y_train, y_test = train_test_split(
        X, y, test_size=test_size, random_state=random_state, stratify=y
    )

    # Scale features
    scaler = StandardScaler()
    X_train_scaled = scaler.fit_transform(X_train)
    X_test_scaled = scaler.transform(X_test)

    return X_train_scaled, X_test_scaled, y_train, y_test, feature_cols, scaler

def prepare_data_neural_only(n_mice=5, n_trials_per_mouse=40, test_size=0.2, random_state=42):
    """
    Prepare data using ONLY neural features (no behavioral/trial info).
    This is the purest test of whether neural activity predicts behavior.

    Returns:
        X_train_scaled: Scaled training features (ONLY neural)
        X_test_scaled: Scaled test features
        y_train: Training labels
        y_test: Test labels
        feature_cols: List of feature column names
        scaler: Fitted StandardScaler object
    """

    # Generate data
    feature_matrix, trial_labels = create_feature_matrix(n_mice, n_trials_per_mouse, random_state=random_state)

    # Create binary target
    feature_matrix['Correct'] = (feature_matrix['Hit'] + feature_matrix['CR']).astype(int)

    # Select only neural features
    feature_cols = ['theta_frequency', 'beta_frequency', 'gamma_frequency', 'noise_level']

    X = feature_matrix[feature_cols]
    y = feature_matrix['Correct']

    # Train-test split
    X_train, X_test, y_train, y_test = train_test_split(
        X, y, test_size=test_size, random_state=random_state, stratify=y
    )

    # Scale features
    scaler = StandardScaler()
    X_train_scaled = scaler.fit_transform(X_train)
    X_test_scaled = scaler.transform(X_test)

    return X_train_scaled, X_test_scaled, y_train, y_test, feature_cols, scaler