Skip to content

Fix: Escape backslashes in additional docstring math expressions #577

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

Open
evasafi wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Fix: Escape backslashes in additional docstring math expressions #577

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
c1551c1
Add Binder configuration for notebooks
invalid-email-address Sep 11, 2026
92a9926
Fix: Escape backslashes in additional docstring math expressions
evasafi Nov 2, 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
22 changes: 22 additions & 0 deletions .binder/environment.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,22 @@
name: aif360-binder
channels:
- conda-forge
- defaults
dependencies:
- python=3.11
- pip
- numpy
- pandas
- scikit-learn
- matplotlib
- seaborn
- jupyter
- ipywidgets
- pip:
- aif360
- cvxpy
- lime
- adversarial-robustness-toolbox>=1.0.0
- tqdm
- lightgbm==3.1.1
- igraph==0.9.8
1 change: 1 addition & 0 deletions .binder/runtime.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
python-3.11
2 changes: 1 addition & 1 deletion aif360/algorithms/preprocessing/optim_preproc_helpers/opt_tools.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -304,7 +304,7 @@ def optimize(self, epsilon=1., dlist=[], verbose=True):
3. Discrimination constraints for all combinations of groups specified
(there is no distinction between protected and unprotected groups).
The constraints are given in eqn 2, 3 in the paper. We use a single
/\epsilon value for all combinations of y and d values
/\\epsilon value for all combinations of y and d values

See section 4.3 in supplementary material of the paper for an example

Expand Down
10 changes: 5 additions & 5 deletions aif360/sklearn/metrics/metrics.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -842,8 +842,8 @@ def smoothed_edf(y_true, y_pred=None, *, prot_attr=None, pos_label=1,
r"""Smoothed empirical differential fairness (EDF).

.. math::
e^{-\epsilon} \leq \frac{\sum_{A=s_i}{P(y|x)} + \alpha}{N_{s_i} + |R_Y|\alpha}
\frac{N_{s_j} + |R_Y|\alpha}{\sum_{A=s_j}{P(y|x) + \alpha}} \leq e^\epsilon
e^{-\\epsilon} \leq \frac{\sum_{A=s_i}{P(y|x)} + \alpha}{N_{s_i} + |R_Y|\alpha}
\frac{N_{s_j} + |R_Y|\alpha}{\sum_{A=s_j}{P(y|x) + \alpha}} \leq e^\\epsilon

See [#foulds18]_ for more details.

Expand All @@ -865,7 +865,7 @@ def smoothed_edf(y_true, y_pred=None, *, prot_attr=None, pos_label=1,
sample_weight (array-like, optional): Sample weights.

Returns:
float: Smoothed EDF, :math:`\epsilon`. Lower is better.
float: Smoothed EDF, :math:`\\epsilon`. Lower is better.

See also:
:func:`intersection`, :func:`smoothed_base_rate`
Expand Down Expand Up @@ -905,8 +905,8 @@ def df_bias_amplification(y_true, y_pred, *, prot_attr=None, pos_label=1,

Returns:
float: Difference in smoothed EDF between the classifier and the
original dataset, :math:`\epsilon_{\text{classifier}}
- \epsilon_{\text{data}}`. Lower is better.
original dataset, :math:`\\epsilon_{\text{classifier}}
- \\epsilon_{\text{data}}`. Lower is better.

References:
.. [#foulds18] J. R. Foulds, R. Islam, K. N. Keya, and S. Pan,
Expand Down