Automate Your ML Pipeline.
Benchmark in One Line.

import pandas as pd
from sklearn.datasets import fetch_california_housing
from multimodel_analysis import MultiModelRegressor

# 1. Load continuous target dataset & prepare feature matrix
data = fetch_california_housing()
X = pd.DataFrame(data.data, columns=data.feature_names).iloc[:500]  # Subset for fast execution
y = data.target[:500]

# 2. Instantiate regression pipeline with scaling
reg = MultiModelRegressor(
    X=X, y=y, test_size=0.2, scaled_data=True, random_state=42
)

# 3. Train & evaluate all 7 built-in regression models
results = reg.run_all_models(random_state=42)

# 4. Display tabular evaluation leaderboard
reg.show_tabular_report()

# 5. Export summary reports & diagnostic scatter plots
reg.get_summary(save_prefix="california_housing", show_plot=False)

import pandas as pd
from xgboost import XGBClassifier
from lightgbm import LGBMClassifier
from multimodel_analysis import MultiModelClassifier, save_report

# 1. Prepare dataset for multi-estimator benchmarking
df = pd.read_csv("credit_risk.csv")
X, y = df.drop(columns=["Default"]), df["Default"]

# 2. Instantiate core MultiModelClassifier instance
clf = MultiModelClassifier(X=X, y=y, test_size=0.20, scaled_data=True)

# 3. Define custom third-party gradient boosting models
custom_models = {
    "XGBoost Classifier": XGBClassifier(n_estimators=200, learning_rate=0.05, random_state=42),
    "LightGBM Classifier": LGBMClassifier(n_estimators=200, learning_rate=0.05, random_state=42)
}

# 4. Benchmark built-in estimators ALONGSIDE XGBoost & LightGBM!
results = clf.run_all_models(custom_models=custom_models)

# 5. Export unified leaderboard comparing standard & boosting models
df_report = clf.show_tabular_report(results, return_df=True)
save_report(df_report, "custom_boosting_benchmark.csv")
clf.plot_comparison(results, save_path="custom_model_comparison.png", show_plot=False)

import pandas as pd
from multimodel_analysis import MultiModelClassifier, save_report

# 1. Execute benchmarking suite to produce evaluation results
df = pd.read_csv("medical_diagnosis.csv")
clf = MultiModelClassifier(X=df.drop(columns=["Target"]), y=df["Target"])
results = clf.run_all_models()

# 2. Capture tabular evaluation metrics DataFrame
df_report = clf.show_tabular_report(results, return_df=True)

# 3. Export to multiple publication & pipeline formats in 1 line each!
save_report(df_report, "exports/benchmark_results.csv")   # CSV Spreadsheet
save_report(df_report, "exports/benchmark_results.xlsx")  # Excel Workbook
save_report(df_report, "exports/benchmark_results.html")  # Styled Web Table
save_report(df_report, "exports/benchmark_results.json")  # REST API JSON Payload

# 4. Generate all diagnostic charts directly into PNG image files
clf.plot_confusion_matrices(results, save_path="exports/confusion_matrix.png", show_plot=False)
clf.plot_roc_curves(results, save_path="exports/roc_curves.png", show_plot=False)
clf.plot_comparison(results, save_path="exports/metrics_summary.png", show_plot=False)