Skip to content

Fix typos and remove unused parameter in ParallelAnalysisNode #105

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

Closed
lucamelis wants to merge 1 commit into from
Closed

Fix typos and remove unused parameter in ParallelAnalysisNode #105

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
11 changes: 5 additions & 6 deletions privacy_guard/analysis/mia/parallel_analysis_node.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,7 +73,7 @@ def __init__(
self._eps_computation_tasks_num = eps_computation_tasks_num

def _compute_metrics_array(
self, args: tuple[str, str, float, int]
self, args: tuple[str, str, int]
) -> list[tuple[float, float, list[float], list[float], list[float]]]:
"""
Make a tuple with two lists:
Expand All @@ -84,12 +84,12 @@ def _compute_metrics_array(
-- A list of epsilons at error thresholds

Args:
A tuple of train/test filenames, delta, and chunk size (number of bootstrap resampling times)
A tuple of train/test filenames and chunk size (number of bootstrap resampling times)
Returns:
A List with elements (accuracy, auc_value, eps_fpr_array, eps_tpr_array, eps_max_array)
"""

train_filename, test_filename, delta, chunk_size = args
train_filename, test_filename, chunk_size = args
loss_train: torch.Tensor = torch.load(train_filename, weights_only=True)
loss_test: torch.Tensor = torch.load(test_filename, weights_only=True)
train_size, test_size = loss_train.shape[0], loss_test.shape[0]
Expand Down Expand Up @@ -148,7 +148,7 @@ def run_analysis(self) -> AnalysisNodeOutput:
Computes analysis outputs based on the input dataframes.
Overrides "BaseAnalysisNode::run_analysis"

First, makes loss_train and loss_test and computes one off metrics like psilon confidence intervals.
First, makes loss_train and loss_test and computes one off metrics like epsilon confidence intervals.

Then, uses "make_metrics_array" to build lists of
metrics, each computed from random subsets of
Expand All @@ -164,7 +164,7 @@ def run_analysis(self) -> AnalysisNodeOutput:
"eps_fpr_max_lb", "eps_fpr_lb", "eps_fpr_ub": epsilon at various false positive rate:
"eps_tpr_lb", "eps_tpr_ub": epsilon at various true positive rates
"eps_max_lb", "eps_max_ub": max of epsilon at various true positive rates and false positive rates
"eps_ci": epsilon calculate with Clopper-Pearson confidence interval
"eps_cp": epsilon calculated with Clopper-Pearson confidence interval
"accuracy", "accuracy_ci": accuracy values
"auc", "auc_ci": area under ROC curve values
"data_size": dictionary with keys "train_size", "test_size", "bootstrap_size"
Expand Down Expand Up @@ -198,7 +198,6 @@ def run_analysis(self) -> AnalysisNodeOutput:
(
train_filename,
test_filename,
self._delta,
chunk_size,
)
for chunk_size in chunk_sizes
Expand Down
Loading