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

[ENH] Implementing Adorn Functions #1439

Open
wants to merge 8 commits into
base: dev
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
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
1 change: 1 addition & 0 deletions AUTHORS.md
Original file line number Diff line number Diff line change
Expand Up @@ -114,3 +114,4 @@ Contributors
- [@joranbeasley](https://github.com/joranbeasley) | [contributions](https://github.com/pyjanitor-devs/pyjanitor/issues?q=is%3Aclosed+mentions%joranbeasley)
-[@kianmeng](https://github.com/kianmeng) | [contributions](https://github.com/pyjanitor-devs/pyjanitor/pull/1290#issue-1906020324)
- [@lbeltrame](https://github.com/lbeltrame) | [contributions](https://github.com/pyjanitor-devs/pyjanitor/pull/1401)
- [@Sarra99](https://github.com/Sarra99) | [contributions](https://github.com/pyjanitor-devs/pyjanitor/issues?q=is%3Aclosed+mentions%3ASarra99)
251 changes: 251 additions & 0 deletions janitor/functions/adorn.py
Copy link
Member

Choose a reason for hiding this comment

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

@Sarra99 would you be kind enough to decorate the functions here as a dataframe method, like we do in other files?

Original file line number Diff line number Diff line change
@@ -0,0 +1,251 @@
from typing import Optional

import pandas as pd


def tabyl(
df: pd.DataFrame,
col1: str,
col2: Optional[str] = None,
col3: Optional[str] = None,
show_counts: bool = True,
show_percentages: bool = False,
percentage_axis: Optional[str] = None, # 'row', 'col', or 'all'
) -> pd.DataFrame:
"""
Create a summary table similar to R's `tabyl`.

Args:
df: Input DataFrame.
col1: Name of the first column for grouping (required).
col2: Name of the second column for grouping (optional).
col3: Name of the third column for grouping (optional).
show_counts: Whether to show raw counts in the table.
show_percentages: Whether to show percentages in the table.
percentage_axis: Axis for percentages ('row', 'col', or 'all').
Only applies if `show_percentages` is True.

Returns:
A DataFrame representing the summary table.

Example :
>>> data = {
... "Category": ["A", "A", "B", "B", "C", "C", "A", "B", "C", "A"],
... "Subcategory": ["X", "Y", "X", "Y", "X", "Y", "X", "Y", "X", "X"],
... "Region": ["North", "South", "East", "West", "North",
... "South", "East", "West", "North", "East"],
... "Value": [1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
... }
>>> df = pd.DataFrame(data)

>>> result = tabyl(df, "Category", "Subcategory", show_percentages=True,
... percentage_axis="row")
>>> print(result)
Subcategory Category X Y
0 A 3.0 (75.00%) 1.0 (25.00%)
1 B 1.0 (33.33%) 2.0 (66.67%)
2 C 2.0 (66.67%) 1.0 (33.33%)

>>> result = tabyl(df, "Category", "Subcategory",
... show_percentages=True, percentage_axis="col")
>>> print(result)
Subcategory Category X Y
0 A 3.0 (50.00%) 1.0 (25.00%)
1 B 1.0 (16.67%) 2.0 (50.00%)
2 C 2.0 (33.33%) 1.0 (25.00%)

"""

if col1 not in df.columns:
raise ValueError(f"Column '{col1}' is not in the DataFrame.")
if col2 and col2 not in df.columns:
raise ValueError(f"Column '{col2}' is not in the DataFrame.")
if col3 and col3 not in df.columns:
raise ValueError(f"Column '{col3}' is not in the DataFrame.")

# Step 1: Group and count
group_cols = [col1]
if col2:
group_cols.append(col2)
if col3:
group_cols.append(col3)

grouped = df.groupby(group_cols).size().reset_index(name="count")

# Step 2: Pivot for 3D (col1, col2, col3)
if col2 and col3:
pivot = grouped.pivot_table(
index=col1,
columns=[col2, col3], # Creating 2-level columns for col2 and col3
values="count",
aggfunc="sum",
fill_value=0,
)
elif col2:
pivot = grouped.pivot_table(
index=col1,
columns=col2,
values="count",
aggfunc="sum",
fill_value=0,
)
else:
pivot = grouped.set_index(col1)["count"].to_frame()

if show_percentages:
pivot = pivot.astype(
float
) # Convert to float before calculating percentages

if percentage_axis == "row":
percentages = pivot.div(pivot.sum(axis=1), axis=0)
elif percentage_axis == "col":
percentages = pivot.div(pivot.sum(axis=0), axis=1)
elif percentage_axis == "all":
total = pivot.values.sum()
percentages = pivot / total
else:
raise ValueError(
"`percentage_axis` must be one of 'row', 'col', or 'all'."
)

percentages = percentages.applymap(lambda x: f"{x:.2%}")

if show_counts:
pivot = pivot.astype(str) + " (" + percentages + ")"
else:
pivot = percentages

return pivot.reset_index()


def adorn_totals(df, col1, col2, axis=0):
"""
Adds a 'Total' row or column to a crosstab generated by tabyl.

:param df: DataFrame used to generate the crosstab
:param col1: First column to create the crosstab
:param col2: Second column to create the crosstab
:param axis: 0 to add a 'Total' row, 1 to add a 'Total' column
:return: DataFrame with a 'Total' row/column added

Example:
>>> data = {
... "Category": ["A", "B", "A", "B", "A", "B", "A", "B"],
... "Subcategory": ["X", "X", "Y", "Y", "X", "X", "Y", "Y"],
... "Value": [1, 2, 3, 4, 5, 6, 7, 8],
... }
>>> df = pd.DataFrame(data)

>>> result = adorn_totals(df, "Category", "Subcategory", axis=0)
>>> print(result)
Subcategory Category X Y
0 A 2 2
1 B 2 2
Total NaN 4 4

>>> result = adorn_totals(df, "Category", "Subcategory", axis=1)
>>> print(result)
Subcategory Category X Y Total
0 A 2 2 4
1 B 2 2 4

"""

# Generate the crosstab using tabyl with the two specified columns
pivot = tabyl(df, col1, col2)

if pivot.empty: # If the crosstab is empty, return it as-is
return pivot

if axis == 0: # Add a 'Total' row
# Select only numeric columns and compute their sum across rows
total_row = pivot.select_dtypes(include="number").sum(axis=0)
total_row.name = "Total" # Set the name of the total row
# Concatenate the total row to the crosstab
pivot = pd.concat([pivot, total_row.to_frame().T])
elif axis == 1: # Add a 'Total' column
# Select only numeric columns and compute their sum across columns
total_col = pivot.select_dtypes(include="number").sum(axis=1)
pivot["Total"] = total_col # Add the total column to the crosstab
else:
raise ValueError(
"The 'axis' argument must be 0 (to add a row) or 1 (to add a column)"
)

return pivot


def adorn_percentages(
df, col1, col2, axis="row", fmt=True, include_ns=False, decimal_places=1
):
"""
Adds percentages to a crosstab generated by tabyl, with options to format
and include raw counts, and also control the behavior
of adorn_pct_formatting and adorn_ns.

:param df: DataFrame used to generate the crosstab
:param col1: First column to create the crosstab
:param col2: Second column to create the crosstab
:param axis: 'row' to add percentages by row, 'col' for column percentages,
'all' for global percentages
:param fmt: If True, formats percentages as strings
(e.g., "12.5%"), else returns numeric values.
:param include_ns: If True, includes raw counts alongside percentages.
:param decimal_places: Number of decimal places for the percentages
:param thousand_separator: Whether to add a thousand separator to the counts
:param percent_format: Whether to format as percentages
:return: DataFrame with percentages and optional formatting and raw counts

"""
# Generate the crosstab using tabyl with the two specified columns
pivot = pd.pivot_table(
df,
values="Value",
index=col1,
columns=col2,
aggfunc="sum",
fill_value=0,
)

if pivot.empty: # If the crosstab is empty, return it as-is
return pivot

# Separate numeric columns from the rest of the data
numeric_cols = pivot.select_dtypes(include="number")

# Calculate the percentages based on the axis
if axis == "row":
percentages = numeric_cols.div(numeric_cols.sum(axis=1), axis=0)
elif axis == "col":
percentages = numeric_cols.div(numeric_cols.sum(axis=0), axis=1)
elif axis == "all":
total_sum = numeric_cols.sum().sum()
percentages = numeric_cols / total_sum
else:
raise ValueError("The 'axis' argument must be 'row', 'col', or 'all'.")

# Format the percentages if requested
if fmt:
percentages = percentages.applymap(
lambda x: f"{x * 100:.{decimal_places}f}%" if pd.notnull(x) else x
)
else:
percentages = percentages.applymap(
lambda x: f"{x:.{decimal_places}f}" if pd.notnull(x) else x
)

# Combine percentages with raw counts if requested (adorn_ns functionality)
if include_ns:
raw_counts = numeric_cols
percentages_with_ns = (
percentages.astype(str) + " (" + raw_counts.astype(str) + ")"
if fmt
else percentages.astype(str) + " (" + raw_counts.astype(str) + ")"
)
percentages = percentages_with_ns

# Reattach the categories and the percentages to form the final DataFrame
result = pd.concat([pivot.iloc[:, :1], percentages], axis=1)

return result
Loading
Loading