From 6b2f8e345a9db4a268868d5554686b5a98018abd Mon Sep 17 00:00:00 2001
From: Valentin Laurent <valentin.laurent.fr@gmail.com>
Date: Mon, 27 Jan 2025 16:26:10 +0100
Subject: [PATCH] ENH: refine classification API
---
mapie_v1/classification.py | 53 ++++++++++----------------------------
1 file changed, 14 insertions(+), 39 deletions(-)
diff --git a/mapie_v1/classification.py b/mapie_v1/classification.py
index 183c37ac..710f27c4 100644
--- a/mapie_v1/classification.py
+++ b/mapie_v1/classification.py
@@ -1,6 +1,6 @@
from __future__ import annotations
-from typing import Optional, Union, List
+from typing import Optional, Union, Tuple, Iterable
from typing_extensions import Self
import numpy as np
@@ -16,7 +16,7 @@ class SplitConformalClassifier:
def __init__(
self,
estimator: ClassifierMixin = LogisticRegression(),
- confidence_level: Union[float, List[float]] = 0.9,
+ confidence_level: Union[float, Iterable[float]] = 0.9,
conformity_score: Union[str, BaseClassificationScore] = "lac",
prefit: bool = True,
n_jobs: Optional[int] = None,
@@ -42,44 +42,31 @@ def conformalize(
return self
def predict(self, X: ArrayLike) -> NDArray:
- """
- Return
- -----
- Return ponctual prediction similar to predict method of
- scikit-learn classifiers
- Shape (n_samples,)
- """
return np.ndarray(0)
- def predict_sets(
+ def predict_set(
self,
X: ArrayLike,
conformity_score_params: Optional[dict] = None,
# Parameters specific to conformal method,
# For example: include_last_label
- ) -> NDArray:
+ ) -> Tuple[NDArray, NDArray]:
"""
- Return
- -----
- An array containing the prediction sets
- Shape (n_samples, n_classes) if confidence_level is float,
- Shape (n_samples, n_classes, confidence_level) if confidence_level
- is a list of floats
+ Shape: (n, ), (n, n_class, n_confidence_level)
"""
- return np.ndarray(0)
+ return np.ndarray(0), np.ndarray(0)
class CrossConformalClassifier:
def __init__(
self,
estimator: ClassifierMixin = LogisticRegression(),
- confidence_level: Union[float, List[float]] = 0.9,
+ confidence_level: Union[float, Iterable[float]] = 0.9,
conformity_score: Union[str, BaseClassificationScore] = "lac",
- cross_val: Union[BaseCrossValidator, str] = 5,
+ cv: Union[int, BaseCrossValidator] = 5,
n_jobs: Optional[int] = None,
verbose: int = 0,
random_state: Optional[Union[int, np.random.RandomState]] = None,
-
) -> None:
pass
@@ -95,34 +82,22 @@ def conformalize(
self,
X_conformalize: ArrayLike,
y_conformalize: ArrayLike,
+ groups: Optional[ArrayLike] = None,
predict_params: Optional[dict] = None
) -> Self:
return self
- def predict(self,
- X: ArrayLike) -> NDArray:
- """
- Return
- -----
- Return ponctual prediction similar to predict method of
- scikit-learn classifiers
- Shape (n_samples,)
- """
+ def predict(self, X: ArrayLike) -> NDArray:
return np.ndarray(0)
- def predict_sets(
+ def predict_set(
self,
X: ArrayLike,
aggregation_method: Optional[str] = "mean",
# How to aggregate the scores by the estimators on test data
conformity_score_params: Optional[dict] = None
- ) -> NDArray:
+ ) -> Tuple[NDArray, NDArray]:
"""
- Return
- -----
- An array containing the prediction sets
- Shape (n_samples, n_classes) if confidence_level is float,
- Shape (n_samples, n_classes, confidence_level) if confidence_level
- is a list of floats
+ Shape: (n, ), (n, n_class, n_confidence_level)
"""
- return np.ndarray(0)
+ return np.ndarray(0), np.ndarray(0)