Skip to content

Adding more metrics to evaluation #149

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
sfluegel05 merged 6 commits into from
Feb 16, 2026
Merged

Adding more metrics to evaluation #149

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
7876520
add senstivity and specificty to evaluation function
schnamo Jan 23, 2026
8f760c9
fixes
schnamo Feb 12, 2026
95d92e5
Merge branch 'dev' of https://github.com/schnamo/python-chebai into dev
schnamo Feb 12, 2026
468d4f4
Merge branch 'ChEB-AI:dev' into dev
schnamo Feb 12, 2026
353ff3c
remove duplicate class
schnamo Feb 13, 2026
ce459ef
Merge branch 'dev' of https://github.com/schnamo/python-chebai into dev
schnamo Feb 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions chebai/preprocessing/bin/smiles_token/tokens.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4373,3 +4373,5 @@ b
[CaH2]
[NH3]
[OH2]
[TlH2+]
[SbH6+3]
48 changes: 40 additions & 8 deletions chebai/result/classification.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,15 +7,18 @@
import torch
from torch import Tensor
from torchmetrics.classification import (
MultilabelF1Score,
MultilabelPrecision,
MultilabelRecall,
MultilabelAUROC,
BinaryF1Score,
BinaryAUROC,
BinaryAveragePrecision,
BinaryF1Score,
BinaryRecall,
MultilabelAUROC,
MultilabelAveragePrecision,
MultilabelF1Score,
MultilabelPrecision,
MultilabelRecall,
MultilabelSpecificity,
)
from torchmetrics.functional import specificity

from chebai.callbacks.epoch_metrics import BalancedAccuracy, MacroF1

Expand Down Expand Up @@ -130,13 +133,39 @@ def metrics_classification_multilabel(
f1_micro = MacroF1(preds.shape[1]).to(device=device)
my_auc_roc = MultilabelAUROC(preds.shape[1]).to(device=device)
my_av_prec = MultilabelAveragePrecision(preds.shape[1]).to(device=device)
my_macro_specificity = MultilabelSpecificity(preds.shape[1], average="macro").to(
device=device
)
my_micro_specificity = MultilabelSpecificity(preds.shape[1], average="micro").to(
device=device
)
my_macro_sensitivity = MultilabelRecall(preds.shape[1], average="macro").to(
device=device
)
my_micro_sensitivity = MultilabelRecall(preds.shape[1], average="micro").to(
device=device
)

macro_f1 = my_f1_macro(preds, labels).cpu().numpy()
micro_f1 = f1_micro(preds, labels).cpu().numpy()
auc_roc = my_auc_roc(preds, labels).cpu().numpy()
prc_auc = my_av_prec(preds, labels).cpu().numpy()

return auc_roc, macro_f1, micro_f1, bal_acc, prc_auc
specificity_macro = my_macro_specificity(preds, labels).cpu().numpy()
specificity_micro = my_micro_specificity(preds, labels).cpu().numpy()
sensitivity_macro = my_macro_sensitivity(preds, labels).cpu().numpy()
sensitivity_micro = my_micro_sensitivity(preds, labels).cpu().numpy()

return (
auc_roc,
macro_f1,
micro_f1,
bal_acc,
prc_auc,
sensitivity_macro,
sensitivity_micro,
specificity_macro,
specificity_micro,
)


def metrics_classification_binary(
Expand All @@ -151,12 +180,15 @@ def metrics_classification_binary(
my_f1 = BinaryF1Score().to(device=device)
my_av_prec = BinaryAveragePrecision().to(device=device)
my_bal_acc = BalancedAccuracy(preds.shape[1]).to(device=device)
my_sensitivity = BinaryRecall().to(device=device)

bal_acc = my_bal_acc(preds, labels).cpu().numpy()
auc_roc = my_auc_roc(preds, labels).cpu().numpy()
# my_auc_roc.update(preds.cpu()[:, 0], labels.cpu()[:, 0])
# auc_roc = my_auc_roc.compute().numpy()
f1_score = my_f1(preds, labels).cpu().numpy()
prc_auc = my_av_prec(preds, labels).cpu().numpy()
sensitivity = my_sensitivity(preds, labels).cpu().numpy()
specificity_result = specificity(preds, labels, task="binary").cpu().numpy()

return auc_roc, f1_score, bal_acc, prc_auc
return auc_roc, f1_score, bal_acc, prc_auc, sensitivity, specificity_result