polartableone

A polars-based implementation of the tableone library for generating summary statistics tables, often referred to as a 'Table 1'.

Features

• Fast summary statistics for a 'Table 1' using polars
• Automatic detection of categorical, continuous, and nonnormal variables
• Support for stratified tables and p-values
• Beautiful table formatting via Great Tables
• Export to HTML, CSV, Excel, LaTeX, and many tabular formats

Differences with tableone Library

• polars rather than pandas
• Use of great_tables to display nicer tables.
• In some cases, use of polars operations rather than numpy, though these are almost always identical.
• Automatic determination of nonnormal columns and will display q1, median, q3 (whereas tableone defaults to showing mean even in these cases).
• Specify number of bins for histograms using show_histograms=8; if show_histograms is set to True, the default is 10 bins (tableone defaults to 8)

But I Prefer tableone

Great! Just convert your polars dataframe to pandas (df.to_pandas()) and then use tableone!

Installation

pip install polartableone

Usage

import polars as pl
from polartableone import TableOne

# load your data
df = pl.read_csv('tests/datasets/pn2012.csv')

# columns to summarize
columns = ['Age', 'SysABP', 'Height', 'Weight', 'ICU', 'death']
# columns containing categorical variables
categorical = ['ICU']
# columns containing categorical variables
continuous = ['Age', 'SysABP', 'Height', 'Weight']
# non-normal variables
nonnormal = ['Age']
# limit the binary variable "death" to a single row
limit = {"death": 1}
# set the order of the categorical variables
order = {"ICU": ["MICU", "SICU", "CSRU", "CCU"]}
# set decimal places for age to 0
decimals = {"Age": 0}
# optionally, a categorical variable for stratification
groupby = 'death'
# rename the death column
labels = {'death': 'Mortality'}

# display minimum and maximum for listed variables
min_max = ['Height']

table = TableOne(df, columns=columns, categorical=categorical, continuous=continuous,
                  groupby=groupby, nonnormal=nonnormal, rename=labels, label_suffix=True,
                  decimals=decimals, limit=limit, min_max=min_max, show_histograms=True,
                  dip_test=True, normal_test=True, tukey_test=True)


# Export to CSV
table.write_csv('table1.csv')

# Export to Great Tables object for further styling
gt_table = table.to_gt()
gt_table.show()