jeremylongshore/tons-of-skills-marketplace / plugins/ai-ml/regression-analysis-tool/skills/performing-regression-analysis
performing-regression-analysis
VerifiedExecute this skill empowers AI assistant to perform regression analysis and modeling using the regression-analysis-tool plugin. it analyzes datasets, generates appropriate regression models (linear, polynomial, etc.), validates the models, and provides performa... Use when analyzing code or data. Trigger with phrases like 'analyze', 'review', or 'examine'.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.7 KB |
| assets/README.md | markdown | 270 B |
| assets/example_dataset.csv | other | 780 B |
| assets/model_visualization_template.py | script | 5.3 KB |
| references/README.md | markdown | 67 B |
| scripts/README.md | markdown | 506 B |
| scripts/evaluate_model.pyexecutable | script | 12.1 KB |
| scripts/generate_model.pyexecutable | script | 12.1 KB |
| scripts/validate_data.pyexecutable | script | 14.4 KB |
scripts/evaluate_model.py
View on GitHub#!/usr/bin/env python3
"""
Evaluate the performance of trained regression models.
This script evaluates regression model performance using metrics such as:
- R-squared (coefficient of determination)
- Mean Squared Error (MSE) and Root Mean Squared Error (RMSE)
- Mean Absolute Error (MAE)
- Mean Absolute Percentage Error (MAPE)
- Residual analysis and diagnostics
"""
import argparse
import sys
import json
import csv
from typing import Dict, List, Tuple, Optional
def load_model(filepath: str) -> Dict:
"""Load a trained model from JSON file."""
try:
with open(filepath, "r", encoding="utf-8") as f:
return json.load(f)
except FileNotFoundError:
print(f"Error: Model file not found: {filepath}", file=sys.stderr)
sys.exit(1)
except json.JSONDecodeError as e:
print(f"Error parsing model JSON: {e}", file=sys.stderr)
sys.exit(1)
def load_csv_data(filepath: str, target_col: str) -> Tuple[List[List[float]], List[float], List[str]]:
"""Load CSV data for evaluation."""
try:
features = []
targets = []
feature_names = []
with open(filepath, "r", encoding="utf-8") as f:
reader = csv.DictReader(f)
columns = reader.fieldnames
if not columns or target_col not in columns:
print(f"Error: Target column '{target_col}' not found", file=sys.stderr)
sys.exit(1)
feature_names = [col for col in columns if col != target_col]
for row in reader:
try:
target = float(row[target_col])
targets.append(target)
feature_values = []
for col in feature_names:
feature_values.append(float(row[col]))
features.append(feature_values)
except (ValueError, KeyError):
pass # Skip invalid rows
if not features or not targets:
print("Error: No valid data could be loaded", file=sys.stderr)
sys.exit(1)
return features, targets, feature_names
except FileNotFoundError:
print(f"Error: File not found: {filepath}", file=sys.stderr)
sys.exit(1)
except Exception as e:
print(f"Error reading file: {e}", file=sys.stderr)
sys.exit(1)
def predict(model: Dict, features: List[List[float]]) -> List[float]:
"""Make predictions with a trained model."""
predictions = []
means = model.get("normalization", {}).get("means", [])
stds = model.get("normalization", {}).get("stds", [])
for row in features:
# Normalize features
normalized_row = row.copy()
if means and stds:
normalized_row = [(row[i] - means[i]) / stds[i] if stds[i] > 0 else 0 for i in range(len(row))]
if model["type"] == "linear":
pred = model["intercept"]
for i, coef in enumerate(model["coefficients"]):
if i < len(normalized_row):
pred += coef * normalized_row[i]
predictions.append(pred)
elif model["type"] == "polynomial":
poly_row = normalized_row.copy()
for d in range(2, model["degree"] + 1):
for val in normalized_row[: len(model["feature_names"])]:
poly_row.append(val**d)
pred = model["intercept"]
for i, coef in enumerate(model["coefficients"]):
if i < len(poly_row):
pred += coef * poly_row[i]
predictions.append(pred)
return predictions
def calculate_metrics(targets: List[float], predictions: List[float]) -> Dict[str, float]:
"""Calculate evaluation metrics."""
n = len(targets)
if n == 0:
return {}
# Basic metrics
mean_target = sum(targets) / n
# Sum of squares
ss_res = sum((targets[i] - predictions[i]) ** 2 for i in range(n))
ss_tot = sum((t - mean_target) ** 2 for t in targets)
# R-squared
r_squared = 1 - (ss_res / ss_tot) if ss_tot > 0 else 0
# MSE and RMSE
mse = ss_res / n
rmse = mse**0.5
# MAE
mae = sum(abs(targets[i] - predictions[i]) for i in range(n)) / n
# MAPE (Mean Absolute Percentage Error)
mape_sum = 0
valid_mape = 0
for i in range(n):
if targets[i] != 0:
mape_sum += abs((targets[i] - predictions[i]) / targets[i])
valid_mape += 1
mape = (mape_sum / valid_mape * 100) if valid_mape > 0 else 0
# Adjusted R-squared (for multiple features)
# adj_r2 = 1 - ((1 - r2) * (n - 1) / (n - k - 1)) where k is number of features
return {
"r_squared": r_squared,
"mse": mse,
"rmse": rmse,
"mae": mae,
"mape": mape,
"mean_target": mean_target,
"std_target": (sum((t - mean_target) ** 2 for t in targets) / n) ** 0.5,
}
def analyze_residuals(targets: List[float], predictions: List[float]) -> Dict:
"""Analyze residuals for model diagnostics."""
residuals = [targets[i] - predictions[i] for i in range(len(targets))]
if not residuals:
return {}
# Residual statistics
mean_residual = sum(residuals) / len(residuals)
std_residual = (sum((r - mean_residual) ** 2 for r in residuals) / len(residuals)) ** 0.5
# Detect systematic bias
positive_residuals = sum(1 for r in residuals if r > 0)
negative_residuals = sum(1 for r in residuals if r < 0)
bias_ratio = max(positive_residuals, negative_residuals) / len(residuals)
# Heteroscedasticity: check if residual variance changes with prediction magnitude
predictions_sorted = sorted(enumerate(predictions), key=lambda x: x[1])
first_half_idx = [i for i, _ in predictions_sorted[: len(predictions) // 2]]
second_half_idx = [i for i, _ in predictions_sorted[len(predictions) // 2 :]]
if first_half_idx and second_half_idx:
var_first = sum((residuals[i] ** 2) for i in first_half_idx) / len(first_half_idx)
var_second = sum((residuals[i] ** 2) for i in second_half_idx) / len(second_half_idx)
heteroscedasticity_score = (
max(var_first, var_second) / min(var_first, var_second) if min(var_first, var_second) > 0 else 1.0
)
else:
heteroscedasticity_score = 1.0
return {
"mean": mean_residual,
"std": std_residual,
"min": min(residuals),
"max": max(residuals),
"positive_count": positive_residuals,
"negative_count": negative_residuals,
"bias_ratio": bias_ratio,
"heteroscedasticity_score": heteroscedasticity_score,
}
def evaluate_model(
model_file: str, data_file: str, target_col: str, output_file: Optional[str] = None, verbose: bool = False
) -> Dict:
"""
Evaluate a trained regression model.
Args:
model_file: Path to trained model JSON
data_file: Path to evaluation data
target_col: Name of target column
output_file: Optional output file
verbose: Verbose output
Returns:
Evaluation results
"""
results = {
"status": "evaluating",
"model_file": model_file,
"data_file": data_file,
"metrics": {},
"residuals": {},
"diagnostics": [],
"evaluation_log": [],
}
# Load model
if verbose:
results["evaluation_log"].append("Loading model...")
model = load_model(model_file)
results["evaluation_log"].append(f"Loaded {model.get('type', 'unknown')} regression model")
# Load evaluation data
if verbose:
results["evaluation_log"].append("Loading evaluation data...")
features, targets, feature_names = load_csv_data(data_file, target_col)
results["evaluation_log"].append(f"Loaded {len(features)} samples for evaluation")
# Make predictions
if verbose:
results["evaluation_log"].append("Making predictions...")
predictions = predict(model, features)
# Calculate metrics
if verbose:
results["evaluation_log"].append("Calculating metrics...")
metrics = calculate_metrics(targets, predictions)
results["metrics"] = metrics
# Analyze residuals
if verbose:
results["evaluation_log"].append("Analyzing residuals...")
residuals = analyze_residuals(targets, predictions)
results["residuals"] = residuals
# Diagnostics
results["diagnostics"] = []
if metrics.get("r_squared", 0) < 0.5:
results["diagnostics"].append("Model has poor explanatory power (R² < 0.5)")
if metrics.get("mape", 0) > 20:
results["diagnostics"].append(f"High Mean Absolute Percentage Error ({metrics['mape']:.1f}%)")
if residuals.get("bias_ratio", 0) > 0.7:
results["diagnostics"].append("Systematic bias detected in predictions")
if residuals.get("heteroscedasticity_score", 1.0) > 3.0:
results["diagnostics"].append("Heteroscedasticity detected (non-constant variance)")
if not results["diagnostics"]:
results["diagnostics"].append("Model performance appears reasonable")
results["status"] = "completed"
# Save results if requested
if output_file:
try:
with open(output_file, "w", encoding="utf-8") as f:
json.dump(results, f, indent=2)
results["evaluation_log"].append(f"Results saved to {output_file}")
except IOError as e:
results["evaluation_log"].append(f"Warning: Could not save results: {e}")
return results
def format_report(results: Dict) -> str:
"""Format evaluation results as a readable report."""
report = []
report.append("=" * 70)
report.append("REGRESSION MODEL EVALUATION REPORT")
report.append("=" * 70)
report.append(f"\nModel: {results['model_file']}")
report.append(f"Data: {results['data_file']}")
metrics = results["metrics"]
report.append("\n[PERFORMANCE METRICS]")
report.append(f" R-squared: {metrics.get('r_squared', 0):.4f}")
report.append(f" RMSE: {metrics.get('rmse', 0):.4f}")
report.append(f" MAE: {metrics.get('mae', 0):.4f}")
report.append(f" MAPE: {metrics.get('mape', 0):.2f}%")
residuals = results["residuals"]
if residuals:
report.append("\n[RESIDUAL ANALYSIS]")
report.append(f" Mean Residual: {residuals.get('mean', 0):.4f}")
report.append(f" Std Residual: {residuals.get('std', 0):.4f}")
report.append(f" Bias Ratio: {residuals.get('bias_ratio', 0):.2f}")
report.append(f" Heteroscedasticity: {residuals.get('heteroscedasticity_score', 0):.2f}")
if results["diagnostics"]:
report.append("\n[DIAGNOSTICS]")
for diag in results["diagnostics"]:
report.append(f" • {diag}")
report.append("\n" + "=" * 70)
return "\n".join(report)
def main():
"""Main entry point."""
parser = argparse.ArgumentParser(
description="Evaluate the performance of trained regression models",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
Examples:
# Evaluate a model on test data
%(prog)s --model model.json --data test_data.csv --target price
# Save evaluation results
%(prog)s --model model.json --data test_data.csv --target price --output eval_results.json
# Verbose output
%(prog)s --model model.json --data test_data.csv --target price --verbose
""",
)
parser.add_argument("-m", "--model", type=str, required=True, help="Path to trained model file (JSON)")
parser.add_argument("-d", "--data", type=str, required=True, help="Path to evaluation data (CSV)")
parser.add_argument("-t", "--target", type=str, required=True, help="Name of target column")
parser.add_argument("-o", "--output", type=str, help="Output file to save evaluation results")
parser.add_argument("-v", "--verbose", action="store_true", help="Enable verbose output")
args = parser.parse_args()
# Evaluate model
results = evaluate_model(args.model, args.data, args.target, args.output, args.verbose)
# Output results
if args.verbose or not args.output:
report = format_report(results)
print(report)
if results["status"] != "completed":
return 1
return 0
if __name__ == "__main__":
sys.exit(main())
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |