ins-pricing 0.1.6__py3-none-any.whl

ins_pricing/modelling/Explain_Run.py

@@ -0,0 +1,36 @@
+from __future__ import annotations
+
+import argparse
+from pathlib import Path
+from typing import Optional
+
+try:
+    from .notebook_utils import run_from_config  # type: ignore
+except Exception:  # pragma: no cover
+    from notebook_utils import run_from_config  # type: ignore
+
+
+def run(config_json: str | Path) -> None:
+    """Run explain by config.json (runner.mode=explain)."""
+    run_from_config(config_json)
+
+
+def _build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="Explain_Run: run explain by config.json (runner.mode=explain)."
+    )
+    parser.add_argument(
+        "--config-json",
+        required=True,
+        help="Path to config.json (relative paths are resolved from ins_pricing/modelling/ when possible).",
+    )
+    return parser
+
+
+def main(argv: Optional[list[str]] = None) -> None:
+    args = _build_parser().parse_args(argv)
+    run(args.config_json)
+
+
+if __name__ == "__main__":
+    main()

ins_pricing/modelling/Explain_entry.py

@@ -0,0 +1,539 @@
+"""Config-driven explain runner for trained BayesOpt models."""
+
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Sequence
+
+import numpy as np
+import pandas as pd
+from sklearn.model_selection import train_test_split
+
+try:
+    from . import bayesopt as ropt  # type: ignore
+    from .cli_common import (  # type: ignore
+        build_model_names,
+        dedupe_preserve_order,
+        load_config_json,
+        normalize_config_paths,
+        resolve_config_path,
+        resolve_path,
+        set_env,
+    )
+except Exception:  # pragma: no cover
+    try:
+        import bayesopt as ropt  # type: ignore
+        from cli_common import (  # type: ignore
+            build_model_names,
+            dedupe_preserve_order,
+            load_config_json,
+            normalize_config_paths,
+            resolve_config_path,
+            resolve_path,
+            set_env,
+        )
+    except Exception:
+        import ins_pricing.bayesopt as ropt  # type: ignore
+        from ins_pricing.cli_common import (  # type: ignore
+            build_model_names,
+            dedupe_preserve_order,
+            load_config_json,
+            normalize_config_paths,
+            resolve_config_path,
+            resolve_path,
+            set_env,
+        )
+
+try:
+    from .run_logging import configure_run_logging  # type: ignore
+except Exception:  # pragma: no cover
+    try:
+        from run_logging import configure_run_logging  # type: ignore
+    except Exception:  # pragma: no cover
+        configure_run_logging = None  # type: ignore
+
+
+_SUPPORTED_METHODS = {"permutation", "shap", "integrated_gradients"}
+_METHOD_ALIASES = {
+    "ig": "integrated_gradients",
+    "integrated": "integrated_gradients",
+    "intgrad": "integrated_gradients",
+}
+
+
+def _safe_name(value: str) -> str:
+    return "".join(ch if ch.isalnum() or ch in "-_." else "_" for ch in str(value))
+
+
+def _load_dataset(path: Path) -> pd.DataFrame:
+    raw = pd.read_csv(path, low_memory=False)
+    raw = raw.copy()
+    for col in raw.columns:
+        s = raw[col]
+        if pd.api.types.is_numeric_dtype(s):
+            raw[col] = pd.to_numeric(s, errors="coerce").fillna(0)
+        else:
+            raw[col] = s.astype("object").fillna("<NA>")
+    return raw
+
+
+def _resolve_path_value(
+    value: Any,
+    *,
+    model_name: str,
+    base_dir: Path,
+    data_dir: Optional[Path] = None,
+) -> Optional[Path]:
+    if value is None:
+        return None
+    if isinstance(value, dict):
+        value = value.get(model_name)
+    if value is None:
+        return None
+    path_str = str(value)
+    try:
+        path_str = path_str.format(model_name=model_name)
+    except Exception:
+        pass
+    if data_dir is not None and not Path(path_str).is_absolute():
+        candidate = data_dir / path_str
+        if candidate.exists():
+            return candidate.resolve()
+    resolved = resolve_path(path_str, base_dir)
+    if resolved is None:
+        return None
+    return resolved
+
+
+def _normalize_methods(raw: Sequence[str]) -> List[str]:
+    methods: List[str] = []
+    for item in raw:
+        key = str(item).strip().lower()
+        if not key:
+            continue
+        key = _METHOD_ALIASES.get(key, key)
+        if key not in _SUPPORTED_METHODS:
+            raise ValueError(f"Unsupported explain method: {item}")
+        methods.append(key)
+    return dedupe_preserve_order(methods)
+
+
+def _save_series(series: pd.Series, path: Path) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    series.to_frame(name="importance").to_csv(path, index=True)
+
+
+def _save_df(df: pd.DataFrame, path: Path) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    df.to_csv(path, index=False)
+
+
+def _shap_importance(values: Any, feature_names: Sequence[str]) -> pd.Series:
+    if isinstance(values, list):
+        values = values[0]
+    arr = np.asarray(values)
+    if arr.ndim == 3:
+        arr = arr[0]
+    scores = np.mean(np.abs(arr), axis=0)
+    return pd.Series(scores, index=list(feature_names)).sort_values(ascending=False)
+
+
+def _parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Run explainability (permutation/SHAP/IG) on trained models."
+    )
+    parser.add_argument(
+        "--config-json",
+        required=True,
+        help="Path to config.json (same schema as training).",
+    )
+    parser.add_argument(
+        "--model-keys",
+        nargs="+",
+        default=None,
+        choices=["glm", "xgb", "resn", "ft", "gnn", "all"],
+        help="Model keys to load for explanation (default from config.explain.model_keys).",
+    )
+    parser.add_argument(
+        "--methods",
+        nargs="+",
+        default=None,
+        help="Explain methods: permutation, shap, integrated_gradients (default from config.explain.methods).",
+    )
+    parser.add_argument(
+        "--output-dir",
+        default=None,
+        help="Override output root for loading models/results.",
+    )
+    parser.add_argument(
+        "--eval-path",
+        default=None,
+        help="Override validation CSV path (supports {model_name}).",
+    )
+    parser.add_argument(
+        "--on-train",
+        action="store_true",
+        help="Explain on train split instead of validation/test.",
+    )
+    parser.add_argument(
+        "--save-dir",
+        default=None,
+        help="Override output directory for explanation artifacts.",
+    )
+    return parser.parse_args()
+
+
+def _explain_for_model(
+    model: ropt.BayesOptModel,
+    *,
+    model_name: str,
+    model_keys: List[str],
+    methods: List[str],
+    on_train: bool,
+    save_dir: Path,
+    explain_cfg: Dict[str, Any],
+) -> None:
+    perm_cfg = dict(explain_cfg.get("permutation") or {})
+    shap_cfg = dict(explain_cfg.get("shap") or {})
+    ig_cfg = dict(explain_cfg.get("integrated_gradients") or {})
+
+    perm_metric = perm_cfg.get("metric", explain_cfg.get("metric", "auto"))
+    perm_repeats = int(perm_cfg.get("n_repeats", 5))
+    perm_max_rows = perm_cfg.get("max_rows", 5000)
+    perm_random_state = perm_cfg.get("random_state", None)
+
+    shap_background = int(shap_cfg.get("n_background", 500))
+    shap_samples = int(shap_cfg.get("n_samples", 200))
+    shap_save_values = bool(shap_cfg.get("save_values", False))
+
+    ig_steps = int(ig_cfg.get("steps", 50))
+    ig_batch_size = int(ig_cfg.get("batch_size", 256))
+    ig_target = ig_cfg.get("target", None)
+    ig_baseline = ig_cfg.get("baseline", None)
+    ig_baseline_num = ig_cfg.get("baseline_num", None)
+    ig_baseline_geo = ig_cfg.get("baseline_geo", None)
+    ig_save_values = bool(ig_cfg.get("save_values", False))
+
+    for key in model_keys:
+        trainer = model.trainers.get(key)
+        if trainer is None:
+            print(f"[Explain] Skip {model_name}/{key}: trainer not available.")
+            continue
+        model.load_model(key)
+        trained_model = getattr(model, f"{key}_best", None)
+        if trained_model is None:
+            print(f"[Explain] Skip {model_name}/{key}: model not loaded.")
+            continue
+
+        if key == "ft" and str(model.config.ft_role) != "model":
+            print(f"[Explain] Skip {model_name}/ft: ft_role != 'model'.")
+            continue
+
+        for method in methods:
+            if method == "permutation" and key not in {"xgb", "resn", "ft"}:
+                print(f"[Explain] Skip permutation for {model_name}/{key}.")
+                continue
+            if method == "shap" and key not in {"glm", "xgb", "resn", "ft"}:
+                print(f"[Explain] Skip shap for {model_name}/{key}.")
+                continue
+            if method == "integrated_gradients" and key not in {"resn", "ft"}:
+                print(f"[Explain] Skip integrated gradients for {model_name}/{key}.")
+                continue
+
+            if method == "permutation":
+                try:
+                    result = model.compute_permutation_importance(
+                        key,
+                        on_train=on_train,
+                        metric=perm_metric,
+                        n_repeats=perm_repeats,
+                        max_rows=perm_max_rows,
+                        random_state=perm_random_state,
+                    )
+                except Exception as exc:
+                    print(f"[Explain] permutation failed for {model_name}/{key}: {exc}")
+                    continue
+                out_path = save_dir / f"{_safe_name(model_name)}_{key}_permutation.csv"
+                _save_df(result, out_path)
+                print(f"[Explain] Saved permutation -> {out_path}")
+
+            if method == "shap":
+                try:
+                    if key == "glm":
+                        shap_result = model.compute_shap_glm(
+                            n_background=shap_background,
+                            n_samples=shap_samples,
+                            on_train=on_train,
+                        )
+                    elif key == "xgb":
+                        shap_result = model.compute_shap_xgb(
+                            n_background=shap_background,
+                            n_samples=shap_samples,
+                            on_train=on_train,
+                        )
+                    elif key == "resn":
+                        shap_result = model.compute_shap_resn(
+                            n_background=shap_background,
+                            n_samples=shap_samples,
+                            on_train=on_train,
+                        )
+                    else:
+                        shap_result = model.compute_shap_ft(
+                            n_background=shap_background,
+                            n_samples=shap_samples,
+                            on_train=on_train,
+                        )
+                except Exception as exc:
+                    print(f"[Explain] shap failed for {model_name}/{key}: {exc}")
+                    continue
+
+                shap_values = shap_result.get("shap_values")
+                X_explain = shap_result.get("X_explain")
+                feature_names = (
+                    list(X_explain.columns)
+                    if isinstance(X_explain, pd.DataFrame)
+                    else list(model.factor_nmes)
+                )
+                importance = _shap_importance(shap_values, feature_names)
+                out_path = save_dir / f"{_safe_name(model_name)}_{key}_shap_importance.csv"
+                _save_series(importance, out_path)
+                print(f"[Explain] Saved SHAP importance -> {out_path}")
+
+                if shap_save_values:
+                    values_path = save_dir / f"{_safe_name(model_name)}_{key}_shap_values.npy"
+                    np.save(values_path, np.array(shap_values, dtype=object), allow_pickle=True)
+                    if isinstance(X_explain, pd.DataFrame):
+                        x_path = save_dir / f"{_safe_name(model_name)}_{key}_shap_X.csv"
+                        _save_df(X_explain, x_path)
+                    meta_path = save_dir / f"{_safe_name(model_name)}_{key}_shap_meta.json"
+                    meta = {
+                        "base_value": shap_result.get("base_value"),
+                        "n_samples": int(len(X_explain)) if X_explain is not None else None,
+                    }
+                    meta_path.write_text(json.dumps(meta, indent=2), encoding="utf-8")
+
+            if method == "integrated_gradients":
+                try:
+                    if key == "resn":
+                        ig_result = model.compute_integrated_gradients_resn(
+                            on_train=on_train,
+                            baseline=ig_baseline,
+                            steps=ig_steps,
+                            batch_size=ig_batch_size,
+                            target=ig_target,
+                        )
+                        series = ig_result.get("importance")
+                        if isinstance(series, pd.Series):
+                            out_path = save_dir / f"{_safe_name(model_name)}_{key}_ig_importance.csv"
+                            _save_series(series, out_path)
+                            print(f"[Explain] Saved IG importance -> {out_path}")
+                        if ig_save_values and "attributions" in ig_result:
+                            attr_path = save_dir / f"{_safe_name(model_name)}_{key}_ig_attributions.npy"
+                            np.save(attr_path, ig_result.get("attributions"))
+                    else:
+                        ig_result = model.compute_integrated_gradients_ft(
+                            on_train=on_train,
+                            baseline_num=ig_baseline_num,
+                            baseline_geo=ig_baseline_geo,
+                            steps=ig_steps,
+                            batch_size=ig_batch_size,
+                            target=ig_target,
+                        )
+                        series_num = ig_result.get("importance_num")
+                        series_geo = ig_result.get("importance_geo")
+                        if isinstance(series_num, pd.Series):
+                            out_path = save_dir / f"{_safe_name(model_name)}_{key}_ig_num_importance.csv"
+                            _save_series(series_num, out_path)
+                            print(f"[Explain] Saved IG num importance -> {out_path}")
+                        if isinstance(series_geo, pd.Series):
+                            out_path = save_dir / f"{_safe_name(model_name)}_{key}_ig_geo_importance.csv"
+                            _save_series(series_geo, out_path)
+                            print(f"[Explain] Saved IG geo importance -> {out_path}")
+                        if ig_save_values:
+                            if ig_result.get("attributions_num") is not None:
+                                attr_path = save_dir / f"{_safe_name(model_name)}_{key}_ig_num_attributions.npy"
+                                np.save(attr_path, ig_result.get("attributions_num"))
+                            if ig_result.get("attributions_geo") is not None:
+                                attr_path = save_dir / f"{_safe_name(model_name)}_{key}_ig_geo_attributions.npy"
+                                np.save(attr_path, ig_result.get("attributions_geo"))
+                except Exception as exc:
+                    print(f"[Explain] integrated gradients failed for {model_name}/{key}: {exc}")
+                    continue
+
+
+def explain_from_config(args: argparse.Namespace) -> None:
+    script_dir = Path(__file__).resolve().parent
+    config_path = resolve_config_path(args.config_json, script_dir)
+    cfg = load_config_json(
+        config_path,
+        required_keys=["data_dir", "model_list", "model_categories", "target", "weight"],
+    )
+    cfg = normalize_config_paths(cfg, config_path)
+
+    set_env(cfg.get("env", {}))
+
+    data_dir = Path(cfg["data_dir"])
+    data_dir.mkdir(parents=True, exist_ok=True)
+
+    output_dir = args.output_dir or cfg.get("output_dir")
+    if isinstance(output_dir, str) and output_dir.strip():
+        resolved = resolve_path(output_dir, config_path.parent)
+        if resolved is not None:
+            output_dir = str(resolved)
+
+    prop_test = cfg.get("prop_test", 0.25)
+    rand_seed = cfg.get("rand_seed", 13)
+
+    explain_cfg = dict(cfg.get("explain") or {})
+
+    model_keys = args.model_keys or explain_cfg.get("model_keys") or ["xgb"]
+    if "all" in model_keys:
+        model_keys = ["glm", "xgb", "resn", "ft", "gnn"]
+    model_keys = dedupe_preserve_order([str(x) for x in model_keys])
+
+    method_list = args.methods or explain_cfg.get("methods") or ["permutation"]
+    methods = _normalize_methods([str(x) for x in method_list])
+
+    on_train = bool(args.on_train or explain_cfg.get("on_train", False))
+
+    model_names = build_model_names(cfg["model_list"], cfg["model_categories"])
+    if not model_names:
+        raise ValueError("No model names generated from model_list/model_categories.")
+
+    save_dir_raw = args.save_dir or explain_cfg.get("save_dir")
+    if save_dir_raw:
+        resolved = resolve_path(str(save_dir_raw), config_path.parent)
+        save_root = resolved if resolved is not None else Path(str(save_dir_raw))
+    else:
+        save_root = None
+
+    for model_name in model_names:
+        train_path = _resolve_path_value(
+            explain_cfg.get("train_path"),
+            model_name=model_name,
+            base_dir=config_path.parent,
+            data_dir=data_dir,
+        )
+        if train_path is None:
+            train_path = data_dir / f"{model_name}.csv"
+        if not train_path.exists():
+            raise FileNotFoundError(f"Missing training dataset: {train_path}")
+
+        validation_override = args.eval_path or explain_cfg.get("validation_path") or explain_cfg.get("eval_path")
+        validation_path = _resolve_path_value(
+            validation_override,
+            model_name=model_name,
+            base_dir=config_path.parent,
+            data_dir=data_dir,
+        )
+
+        raw = _load_dataset(train_path)
+        if validation_path is not None:
+            if not validation_path.exists():
+                raise FileNotFoundError(f"Missing validation dataset: {validation_path}")
+            train_df = raw
+            test_df = _load_dataset(validation_path)
+        else:
+            if float(prop_test) <= 0:
+                train_df = raw
+                test_df = raw.copy()
+            else:
+                train_df, test_df = train_test_split(
+                    raw, test_size=prop_test, random_state=rand_seed
+                )
+
+        binary_target = cfg.get("binary_target") or cfg.get("binary_resp_nme")
+        feature_list = cfg.get("feature_list")
+        categorical_features = cfg.get("categorical_features")
+
+        model = ropt.BayesOptModel(
+            train_df,
+            test_df,
+            model_name,
+            cfg["target"],
+            cfg["weight"],
+            feature_list,
+            binary_resp_nme=binary_target,
+            cate_list=categorical_features,
+            prop_test=prop_test,
+            rand_seed=rand_seed,
+            epochs=int(cfg.get("epochs", 50)),
+            use_gpu=bool(cfg.get("use_gpu", True)),
+            output_dir=output_dir,
+            xgb_max_depth_max=int(cfg.get("xgb_max_depth_max", 25)),
+            xgb_n_estimators_max=int(cfg.get("xgb_n_estimators_max", 500)),
+            resn_weight_decay=cfg.get("resn_weight_decay"),
+            final_ensemble=bool(cfg.get("final_ensemble", False)),
+            final_ensemble_k=int(cfg.get("final_ensemble_k", 3)),
+            final_refit=bool(cfg.get("final_refit", True)),
+            optuna_storage=cfg.get("optuna_storage"),
+            optuna_study_prefix=cfg.get("optuna_study_prefix"),
+            best_params_files=cfg.get("best_params_files"),
+            gnn_use_approx_knn=cfg.get("gnn_use_approx_knn", True),
+            gnn_approx_knn_threshold=cfg.get("gnn_approx_knn_threshold", 50000),
+            gnn_graph_cache=cfg.get("gnn_graph_cache"),
+            gnn_max_gpu_knn_nodes=cfg.get("gnn_max_gpu_knn_nodes", 200000),
+            gnn_knn_gpu_mem_ratio=cfg.get("gnn_knn_gpu_mem_ratio", 0.9),
+            gnn_knn_gpu_mem_overhead=cfg.get("gnn_knn_gpu_mem_overhead", 2.0),
+            ft_role=str(cfg.get("ft_role", "model")),
+            ft_feature_prefix=str(cfg.get("ft_feature_prefix", "ft_emb")),
+            ft_num_numeric_tokens=cfg.get("ft_num_numeric_tokens"),
+            infer_categorical_max_unique=int(cfg.get("infer_categorical_max_unique", 50)),
+            infer_categorical_max_ratio=float(cfg.get("infer_categorical_max_ratio", 0.05)),
+            reuse_best_params=bool(cfg.get("reuse_best_params", False)),
+        )
+
+        model_dir_override = _resolve_path_value(
+            explain_cfg.get("model_dir"),
+            model_name=model_name,
+            base_dir=config_path.parent,
+            data_dir=None,
+        )
+        if model_dir_override is not None:
+            model.output_manager.model_dir = model_dir_override
+        result_dir_override = _resolve_path_value(
+            explain_cfg.get("result_dir") or explain_cfg.get("results_dir"),
+            model_name=model_name,
+            base_dir=config_path.parent,
+            data_dir=None,
+        )
+        if result_dir_override is not None:
+            model.output_manager.result_dir = result_dir_override
+        plot_dir_override = _resolve_path_value(
+            explain_cfg.get("plot_dir"),
+            model_name=model_name,
+            base_dir=config_path.parent,
+            data_dir=None,
+        )
+        if plot_dir_override is not None:
+            model.output_manager.plot_dir = plot_dir_override
+
+        if save_root is None:
+            save_dir = Path(model.output_manager.result_dir) / "explain"
+        else:
+            save_dir = Path(save_root)
+        save_dir.mkdir(parents=True, exist_ok=True)
+
+        print(f"\n=== Explain model {model_name} ===")
+        _explain_for_model(
+            model,
+            model_name=model_name,
+            model_keys=model_keys,
+            methods=methods,
+            on_train=on_train,
+            save_dir=save_dir,
+            explain_cfg=explain_cfg,
+        )
+
+
+def main() -> None:
+    if configure_run_logging:
+        configure_run_logging(prefix="explain_entry")
+    args = _parse_args()
+    explain_from_config(args)
+
+
+if __name__ == "__main__":
+    main()

ins_pricing/modelling/Pricing_Run.py

@@ -0,0 +1,36 @@
+from __future__ import annotations
+
+import argparse
+from pathlib import Path
+from typing import Optional
+
+try:
+    from .notebook_utils import run_from_config  # type: ignore
+except Exception:  # pragma: no cover
+    from notebook_utils import run_from_config  # type: ignore
+
+
+def run(config_json: str | Path) -> None:
+    """Unified entry point: run entry/incremental/watchdog/DDP based on config.json runner."""
+    run_from_config(config_json)
+
+
+def _build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="Pricing_Run: run BayesOpt by config.json (entry/incremental/watchdog/DDP)."
+    )
+    parser.add_argument(
+        "--config-json",
+        required=True,
+        help="Path to config.json (relative paths are resolved from ins_pricing/modelling/ when possible).",
+    )
+    return parser
+
+
+def main(argv: Optional[list[str]] = None) -> None:
+    args = _build_parser().parse_args(argv)
+    run(args.config_json)
+
+
+if __name__ == "__main__":
+    main()

ins_pricing/modelling/README.md

@@ -0,0 +1,33 @@
+# ins_pricing
+
+This directory contains reusable production-grade tooling and training frameworks, with a focus on the BayesOpt series.
+
+Key contents:
+- `bayesopt/`: core subpackage (data preprocessing, trainers, models, plotting, explainability)
+- `plotting/`: standalone plotting helpers (lift/roc/importance/geo)
+- `explain/`: explainability helpers (Permutation/Integrated Gradients/SHAP)
+- `BayesOpt.py`: compatibility entry point for legacy imports
+- `BayesOpt_entry.py`: batch training CLI
+- `BayesOpt_incremental.py`: incremental training CLI
+- `cli_common.py` / `notebook_utils.py`: shared CLI and notebook utilities
+- `demo/config_template.json` / `demo/config_incremental_template.json`: config templates
+- `Explain_entry.py` / `Explain_Run.py`: explainability entry points (load trained models)
+- `demo/config_explain_template.json` / `demo/Explain_Run.ipynb`: explainability demo
+
+Note: `modelling/demo/` is kept in the repo only and is not shipped in the PyPI package.
+
+Common usage:
+- CLI: `python ins_pricing/modelling/BayesOpt_entry.py --config-json ...`
+- Notebook: `from ins_pricing.bayesopt import BayesOptModel`
+
+Explainability (load trained models under `Results/model` and explain a validation set):
+- CLI: `python ins_pricing/modelling/Explain_entry.py --config-json ins_pricing/modelling/demo/config_explain_template.json`
+- Notebook: open `ins_pricing/modelling/demo/Explain_Run.ipynb` and run it
+
+Notes:
+- Models load from `output_dir/model` by default (override with `explain.model_dir`).
+- Validation data can be specified via `explain.validation_path`.
+
+Operational notes:
+- Training outputs are written to `plot/`, `Results/`, and `model/` by default.
+- Keep large data and secrets outside the repo and use environment variables or `.env`.

ins_pricing/modelling/__init__.py

@@ -0,0 +1,44 @@
+from __future__ import annotations
+
+from importlib import import_module
+
+# Keep imports lazy to avoid hard dependencies when only using lightweight modules.
+
+__all__ = [
+    "BayesOptConfig",
+    "BayesOptModel",
+    "IOUtils",
+    "TrainingUtils",
+    "free_cuda",
+    "bayesopt",
+    "plotting",
+    "explain",
+]
+
+_LAZY_ATTRS = {
+    "bayesopt": "ins_pricing.modelling.bayesopt",
+    "plotting": "ins_pricing.modelling.plotting",
+    "explain": "ins_pricing.modelling.explain",
+    "BayesOptConfig": "ins_pricing.modelling.bayesopt.core",
+    "BayesOptModel": "ins_pricing.modelling.bayesopt.core",
+    "IOUtils": "ins_pricing.modelling.bayesopt.utils",
+    "TrainingUtils": "ins_pricing.modelling.bayesopt.utils",
+    "free_cuda": "ins_pricing.modelling.bayesopt.utils",
+}
+
+
+def __getattr__(name: str):
+    target = _LAZY_ATTRS.get(name)
+    if not target:
+        raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
+    module = import_module(target)
+    if name in {"bayesopt", "plotting", "explain"}:
+        value = module
+    else:
+        value = getattr(module, name)
+    globals()[name] = value
+    return value
+
+
+def __dir__() -> list[str]:
+    return sorted(set(__all__) | set(globals().keys()))