"""
Conditional Data Imputation
============================

This example shows how to use :class:`~shapiq.ConditionalImputer` for
conditional (observational) imputation when computing Shapley interactions.
Conditional imputation respects feature dependencies, unlike marginal
(interventional) imputation.
"""

from __future__ import annotations

from sklearn.ensemble import RandomForestRegressor
from sklearn.model_selection import train_test_split

import shapiq

# %%
# Load Data and Train Model
# --------------------------

X, y = shapiq.load_california_housing()
X_train, X_test, y_train, y_test = train_test_split(
    X.values,
    y.values,
    test_size=0.25,
    random_state=42,
)
n_features = X_train.shape[1]

model = RandomForestRegressor(
    n_estimators=100,
    max_depth=n_features,
    max_features=2 / 3,
    max_samples=2 / 3,
    random_state=42,
)
model.fit(X_train, y_train)
print(f"Train R2: {model.score(X_train, y_train):.4f}")
print(f"Test  R2: {model.score(X_test, y_test):.4f}")

# %%
# Conditional Imputer
# --------------------
# Set ``imputer="conditional"`` in :class:`~shapiq.TabularExplainer`.
# The imputer trains a gradient boosting model per feature to learn the
# conditional distribution. Key parameters:
#
# - ``sample_size``: samples drawn from conditional background
# - ``conditional_budget``: coalitions per data point for training
# - ``conditional_threshold``: quantile threshold for neighbourhood

explainer = shapiq.TabularExplainer(
    model=model,
    data=X_train,
    index="SII",
    max_order=2,
    imputer="conditional",
    sample_size=100,
    conditional_budget=32,
    conditional_threshold=0.04,
)

# %%
# Explain a Single Instance
# --------------------------

x_explain = X_test[100]
iv = explainer.explain(x_explain, budget=2**n_features, random_state=0)
print(iv)

# %%
# Network Plot
# -------------

shapiq.network_plot(interaction_values=iv, feature_names=list(X.columns))