tvtsplit 0.0.2a1

No project description provided

tvtsplit

---

Split polars DataFrame into training data, validation data and test data.

Table of Contents

• Installation
• License

Installation

pip install tvtsplit

Usage

import polars as pl
import tvtsplit

params = {"shuffle": False, "seed": 0, "val_size": 0.25, "test_size": 0.25}
sample_df = pl.DataFrame(
    data=["aaa", "bbb", "ccc", "ddd", "eee", "fff"],
    schema=[("txt", pl.String)],
    )

print(sample_df.tvtsplit.train(params))
shape: (4, 1)
┌─────┐
│ txt │
│ --- │
│ str │
╞═════╡
│ aaa │
│ bbb │
│ ccc │
│ ddd │
└─────┘

print(sample_df.tvtsplit.val(params))
shape: (1, 1)
┌─────┐
│ txt │
│ --- │
│ str │
╞═════╡
│ eee │
└─────┘

print(sample_df.tvtsplit.test(params))
shape: (1, 1)
┌─────┐
│ txt │
│ --- │
│ str │
╞═════╡
│ fff │
└─────┘

License

tvtsplit is distributed under the terms of the MIT license.