Source code for caliber.binary_classification.ood.da_histogram_binning

import numpy as np
from scipy.stats import expon

from caliber.binary_classification.base import AbstractBinaryClassificationModel


[docs] class DistanceAwareHistogramBinningBinaryClassificationModel( AbstractBinaryClassificationModel ): def __init__( self, n_prob_bins: int = 10, n_dist_bins: int = 10, conf_distance: float = 0.95, min_prob_bin: float = 0.0, ): super().__init__() self.n_prob_bins = n_prob_bins self.n_dist_bins = n_dist_bins self.conf_distance = conf_distance self._min_prob_bin = min_prob_bin self._prob_bin_edges = None self._dist_bin_edges = None self._cdf = None self._max_distance = None self._quantile_distance = None
[docs] def fit(self, probs: np.ndarray, distances: np.ndarray, targets: np.ndarray): self._quantile_distance = np.quantile(distances, self.conf_distance) self._max_distance = np.max(distances) self._cdf = expon(0, self._max_distance).cdf self._prob_bin_edges = self._get_prob_bin_edges() self._dist_bin_edges = self._get_dist_bin_edges() prob_bin_indices = np.digitize(probs, self._prob_bin_edges) dist_bin_indices = np.digitize(distances, self._dist_bin_edges) self._params = np.empty((self.n_prob_bins + 1, self.n_dist_bins + 1)) for i in range(1, self.n_prob_bins + 2): for j in range(1, self.n_dist_bins + 2): mask = self._get_mask(i, j, prob_bin_indices, dist_bin_indices) self._fit_bin(i, j, mask, targets)
[docs] def predict_proba(self, probs: np.ndarray, distances: np.ndarray) -> np.ndarray: if self._prob_bin_edges is None: raise ValueError("Run `fit` first.") prob_bin_indices = np.digitize(probs, self._prob_bin_edges) dist_bin_indices = np.digitize(distances, self._dist_bin_edges) probs = np.copy(probs) for i in range(1, self.n_prob_bins + 2): for j in range(1, self.n_dist_bins + 2): mask = self._get_mask(i, j, prob_bin_indices, dist_bin_indices) if not np.isnan(self._params[i - 1, j - 1]): probs[mask] = self._params[i - 1, j - 1] else: cdf = self._get_cdf(distances[mask]) probs[mask] = (1 - cdf) * probs[mask] + 0.5 * cdf return probs
[docs] def predict(self, probs: np.ndarray, distances: np.ndarray) -> np.ndarray: return (self.predict_proba(probs, distances) >= 0.5).astype(int)
def _get_prob_bin_edges(self) -> np.ndarray: return np.linspace(0, 1, self.n_prob_bins + 1) def _get_dist_bin_edges(self) -> np.ndarray: return np.linspace(0, self._max_distance, self.n_dist_bins + 1) def _fit_bin(self, i: int, j: int, mask: np.ndarray, targets: np.ndarray): prob_bin = np.mean(mask) if prob_bin >= self._min_prob_bin: cdf = self._get_cdf(self._dist_bin_edges[j - 1]) self._params[i - 1, j - 1] = (1 - cdf) * np.mean(targets[mask]) + 0.5 * cdf else: self._params[i - 1, j - 1] = np.nan @staticmethod def _get_mask( prob_bin_idx: int, dist_bin_idx: int, prob_bin_indices: np.ndarray, dist_bin_indices: np.ndarray, ) -> np.ndarray: return (prob_bin_indices == prob_bin_idx) * (dist_bin_indices == dist_bin_idx) def _get_cdf(self, d: float) -> np.ndarray: d -= self._quantile_distance return self._cdf(d) * (d > 0)