File size: 2,448 Bytes
d7e53e8 | 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 | import matplotlib.pyplot as plt
import pandas as pd
from sklearn.model_selection import learning_curve
from sklearn.metrics import (
roc_curve, precision_recall_curve,
confusion_matrix, classification_report
)
def regression_graphs(graph_type, X, y, model, pipeline, y_test, preds):
if graph_type == "Actual vs Predicted":
fig, ax = plt.subplots()
ax.plot(y_test.values[:100])
ax.plot(preds[:100])
ax.legend(["Actual", "Predicted"])
elif graph_type == "Residual Plot":
fig, ax = plt.subplots()
ax.scatter(preds, y_test - preds)
ax.axhline(0)
elif graph_type == "Residual Histogram":
fig, ax = plt.subplots()
ax.hist(y_test - preds, bins=30)
elif graph_type == "Feature Importance":
fig = None
if hasattr(model, "feature_importances_"):
fig, ax = plt.subplots()
ax.bar(range(len(model.feature_importances_)), model.feature_importances_)
elif graph_type == "Learning Curve":
sizes, train_scores, test_scores = learning_curve(
pipeline, X, y
)
fig, ax = plt.subplots()
ax.plot(sizes, train_scores.mean(axis=1))
ax.plot(sizes, test_scores.mean(axis=1))
ax.legend(["Train", "Test"])
return fig
def classification_graphs(graph_type, pipeline, X_test, y_test, preds):
if graph_type == "Confusion Matrix":
cm = confusion_matrix(y_test, preds)
fig, ax = plt.subplots()
ax.imshow(cm)
ax.set_title("Confusion Matrix")
elif graph_type == "ROC Curve":
probs = pipeline.predict_proba(X_test)[:, 1]
fpr, tpr, _ = roc_curve(y_test, probs)
fig, ax = plt.subplots()
ax.plot(fpr, tpr)
ax.set_title("ROC Curve")
elif graph_type == "Per-Class Metrics Table":
fig = classification_report(y_test, preds, output_dict=True)
fig = pd.DataFrame(fig).transpose()
elif graph_type == "Precision-Recall Curve":
probs = pipeline.predict_proba(X_test)[:, 1]
p, r, _ = precision_recall_curve(y_test, probs)
fig, ax = plt.subplots()
ax.plot(r, p)
ax.set_title("Precision-Recall Curve")
elif graph_type == "Probability Histogram":
probs = pipeline.predict_proba(X_test)[:, 1]
fig, ax = plt.subplots()
ax.hist(probs, bins=20)
ax.set_title("Prediction Probability Histogram")
return fig |