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.

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

Jump to bottom

Little refactor and minor bug #41

Open
wants to merge 7 commits into
base: main
Choose a base branch
from
+24 −10
Open

Little refactor and minor bug #41

Show file tree
Hide file tree
Changes from 5 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
da8b860
fix/feat: Improved loading function using pandas only instead of StrinIO
baraldian Nov 25, 2022
8f99ce3
add: mobility filter to speed up filter
baraldian Feb 2, 2023
7474ac6
Merge remote-tracking branch 'origin-folktables/main' into main
baraldian Mar 27, 2023
1762722
Merge remote-tracking branch 'origin-folktables/main'
baraldian Mar 25, 2024
97bfde3
fix: refactor fillna lambda function with a function that also apply …
baraldian Apr 15, 2024
b1d534e
Merge branch 'main' into main
mrtzh May 1, 2024
6689deb
Merge branch 'main' into main
mrtzh May 7, 2024
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
31 changes: 22 additions & 9 deletions folktables/acs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,6 +65,9 @@ def get_definitions(self, download=False):
return load_definitions(root_dir=self._root_dir, year=self._survey_year, horizon=self._horizon,
download=download)

def fillna_safe(x, value=-1):
Copy link
Member

@mrtzh mrtzh May 1, 2024 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you describe what the intended behavior of this function should be?

Note that np.nan_to_num will convert nans to 0.0 unless you set the keyword nan=value. See here.

So, currently, I believe this function will first convert all nans to zero and then pass an array without any nans to pandas.

Am I getting this wrong?

x = np.nan_to_num(x, value)
return pd.DataFrame(x).fillna(value=value).values

def adult_filter(data):
"""Mimic the filters in place for Adult data.
Expand Down Expand Up @@ -98,7 +101,7 @@ def adult_filter(data):
target_transform=lambda x: x > 50000,
group='RAC1P',
preprocess=adult_filter,
postprocess=lambda x: np.nan_to_num(x, -1),
postprocess=fillna_safe,
)

ACSEmployment = folktables.BasicProblem(
Expand All @@ -124,7 +127,7 @@ def adult_filter(data):
target_transform=lambda x: x == 1,
group='RAC1P',
preprocess=lambda x: x,
postprocess=lambda x: np.nan_to_num(x, -1),
postprocess=fillna_safe,
)

ACSHealthInsurance = folktables.BasicProblem(
Expand Down Expand Up @@ -159,7 +162,7 @@ def adult_filter(data):
target_transform=lambda x: x == 1,
group='RAC1P',
preprocess=lambda x: x,
postprocess=lambda x: np.nan_to_num(x, -1),
postprocess=fillna_safe,
)

def public_coverage_filter(data):
Expand Down Expand Up @@ -197,7 +200,7 @@ def public_coverage_filter(data):
target_transform=lambda x: x == 1,
group='RAC1P',
preprocess=public_coverage_filter,
postprocess=lambda x: np.nan_to_num(x, -1),
postprocess=fillna_safe,
)

def travel_time_filter(data):
Expand Down Expand Up @@ -233,9 +236,19 @@ def travel_time_filter(data):
target_transform=lambda x: x > 20,
group='RAC1P',
preprocess=travel_time_filter,
postprocess=lambda x: np.nan_to_num(x, -1),
postprocess=fillna_safe,
)


def mobility_filter(data):
"""
Filters for the employment prediction task
"""
df = data
df = df[df['AGEP'] > 18]
df = df[df['AGEP'] < 35]
return df

ACSMobility = folktables.BasicProblem(
features=[
'AGEP',
Expand Down Expand Up @@ -263,8 +276,8 @@ def travel_time_filter(data):
target="MIG",
target_transform=lambda x: x == 1,
group='RAC1P',
preprocess=lambda x: x.drop(x.loc[(x['AGEP'] <= 18) | (x['AGEP'] >= 35)].index),
postprocess=lambda x: np.nan_to_num(x, -1),
preprocess=mobility_filter,
postprocess=fillna_safe,
)

def employment_filter(data):
Expand Down Expand Up @@ -301,7 +314,7 @@ def employment_filter(data):
target_transform=lambda x: x == 1,
group='RAC1P',
preprocess=employment_filter,
postprocess=lambda x: np.nan_to_num(x, -1),
postprocess=fillna_safe,
)

ACSIncomePovertyRatio = folktables.BasicProblem(
Expand Down Expand Up @@ -331,5 +344,5 @@ def employment_filter(data):
target_transform=lambda x: x < 250,
group='RAC1P',
preprocess=lambda x: x,
postprocess=lambda x: np.nan_to_num(x, -1),
postprocess=fillna_safe,
)
3 changes: 2 additions & 1 deletion folktables/load_acs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -144,6 +144,7 @@ def load_definitions(root_dir, year=2018, horizon='1-Year', download=False):
year_string = year if horizon == '1-Year' else f'{year - 4}-{year}'
url = f'https://www2.census.gov/programs-surveys/acs/tech_docs/pums/data_dict/PUMS_Data_Dictionary_{year_string}.csv'

os.makedirs(base_datadir, exist_ok=True)
response = requests.get(url)
with open(file_path, 'wb') as handle:
handle.write(response.content)
Expand Down Expand Up @@ -191,4 +192,4 @@ def generate_categories(features, definition_df):
del mapping_dict[-99999999999999.0]

categories[feature] = mapping_dict
return categories
return categories
Loading