pytae 2.6.3

A package to enhance common day to day tasks in the life of a data scientist

pytae

pytae is a lightweight Python package that makes everyday data science tasks faster and more readable — think SAS-style chaining on top of pandas.

Installation

pip install pytae

Requires Python ≥ 3.7, pandas ≥ 1.0.0, seaborn ≥ 0.10.0.

Quick start

import pytae as pt

df = pt.sample_data['penguins']

(df
 .qry({'species': ['Adelie', 'Gentoo'], 'sex': 'Female'})
 .select(['species', 'island', 'bill_length_mm', 'body_mass_g'])
 .agg_df(aggfunc=['mean', 'n'])
)

Features

Six feature sets, each with a companion reference notebook:

1) Plotting — Plotter

Lightweight plotting built on top of pandas.plot, fully compatible with matplotlib. plotter.ipynb

2) Filtering — qry()

Filter rows using a dict of conditions: equality, list, comparison operators, interval notation, and not in. qry.ipynb

3) Selection — select()

Pick columns by name, regex, slice, dtype, substring match, or callable. Pin columns up front with everything(). select.ipynb

4) Reshaping — long(), wide()

long() melts all numeric columns to rows. wide() pivots a column's values into headers. Designed to work as a pair. shape.ipynb

5) Aggregation — agg_df()

Auto-detects group columns (non-numeric) and aggregates the rest. Supports str, list, and dict aggfunc. n is a special token for group count. agg_df.ipynb

6) Utilities — clip(), handle_missing(), cols(), group_x()

• cols() — sorted column list
• handle_missing() — fill NaN with a visible marker before aggregating
• group_x() — broadcast a group aggregate back to every row (like transform)
• clip() — copy DataFrame to clipboard for pasting into Excel / Sheets

other_utilities.ipynb