Skip to content
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.

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

numpy.core._exceptions._ArrayMemoryError: Unable to allocate 456. TiB for an array with shape (64834560, 967680) and data type float64 #439

Open
pradeepdev-1995 opened this issue Jan 27, 2023 · 0 comments

Comments

@pradeepdev-1995
Copy link

I tried to mitigate bias using OptimPreproc in my dataset. Using the code below

from aif360.algorithms.preprocessing.optim_preproc import OptimPreproc
from aif360.algorithms.preprocessing.optim_preproc_helpers.opt_tools import OptTools
from aif360.algorithms.preprocessing.optim_preproc_helpers.distortion_functions import get_distortion_adult
from aif360.metrics import ClassificationMetric
from aif360.metrics import BinaryLabelDatasetMetric
from aif360.datasets.multiclass_label_dataset import MulticlassLabelDataset
import pandas as pd
import numpy as np
data = pd.read_csv("Train.csv")
dataset = MulticlassLabelDataset(
                                 favorable_label=[0,3,1],
                                 unfavorable_label = [2],
                                 df=data,
                                 label_names=['Segmentation'],
                                 protected_attribute_names=['Age']
                                )

attr = dataset.protected_attribute_names[0]
idx = dataset.protected_attribute_names.index(attr)
privileged_groups = [{attr: dataset.privileged_protected_attributes[idx][0]}]
unprivileged_groups = [{attr: dataset.unprivileged_protected_attributes[idx][0]}]
metric_pred = BinaryLabelDatasetMetric(dataset,
                                       unprivileged_groups=unprivileged_groups,
                                       privileged_groups=privileged_groups)

def get_distortion_custom(vold, vnew):
    def adjustAge(a):
        if a == '>=70':
            return 70.0
        else:
            return float(a)

    bad_val = 3.0
    aOld = adjustAge(vold['Age'])
    aNew = adjustAge(vnew['Age'])

    # Age cannot be increased or decreased in more than a decade
    if np.abs(aOld-aNew) > 10.0:
        return bad_val

    # Penalty of 2 if age is decreased or increased
    if np.abs(aOld-aNew) > 0:
        return 2.0

#optimized preprocessing bias mitigation
optim_options = {
    "distortion_fun": get_distortion_custom,
    "epsilon": 0.05,
    "clist": [0.99, 1.99, 2.99],
    "dlist": [.1, 0.05, 0]
}

OP = OptimPreproc(OptTools, optim_options)
OP = OP.fit(dataset)
dataset_transf_train = OP.transform(dataset, transform_Y=True)

But it return the error like this

numpy.core._exceptions._ArrayMemoryError: Unable to allocate 456. TiB for an array with shape (64834560, 967680) and data type float64

So how to put our own distortion_fun in OptimPreproc for our own dataset

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant