https://github.com/drhagen/tabeline

User-friendly data frame and data grammar library for Python
https://github.com/drhagen/tabeline

data-grammar data-table dplyr

Last synced: about 2 months ago
JSON representation

User-friendly data frame and data grammar library for Python

• Host: GitHub
• URL: https://github.com/drhagen/tabeline
• Owner: drhagen
• License: mit
• Created: 2022-02-12T20:52:33.000Z (almost 3 years ago)
• Default Branch: master
• Last Pushed: 2024-08-19T18:23:38.000Z (4 months ago)
• Last Synced: 2024-09-18T00:06:10.744Z (3 months ago)
• Topics: data-grammar, data-table, dplyr
• Language: Python
• Homepage: https://tabeline.drhagen.com
• Size: 1.61 MB
• Stars: 14
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Contributing: docs/contributing.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Tabeline

Tabeline is a data frame and data grammar library. You write the expressions in strings and supply them to methods on the `DataFrame` class. The strings are parsed by Parsita and converted into Polars for execution.

Tabeline draws inspiration from dplyr, the data grammar of R's tidyverse, especially for its methods names. The `filter`, `mutate`, `group_by`, and `summarize` methods should all feel familiar. But Tabeline is as proper a Python library as can be, using methods instead of pipes, like is standard in R. 

Tabeline uses Polars under the hood, but adds a lot of handling of edge cases from Polars, which otherwise result in crashes or behavior that is not type stable.

See the [Documentation](https://tabeline.drhagen.com) for the full user guide.

## Installation

It is recommended to install Tabeline from PyPI using `pip`.

```shell

pip install tabeline

```

## Motivating example

```python

from tabeline import DataFrame

# Construct a data frame using clean syntax

# from_csv, from_pandas, and from_polars are also available 

df = DataFrame(

    id=[0, 0, 0, 0, 1, 1, 1, 1, 1],

    t=[0, 6, 12, 24, 0, 6, 12, 24, 48],

    y=[0, 2, 3, 1, 0, 4, 3, 2, 1],

)

# Use data grammar methods and string expressions to define

# transformed data frames

analysis = (

    df

    .filter("t <= 24")

    .group_by("id")

    .summarize(auc="trapz(t, y)")

)

print(analysis)

# shape: (2, 2)

# ┌─────┬──────┐

# │ id  ┆ auc  │

# │ --- ┆ ---  │

# │ i64 ┆ f64  │

# ╞═════╪══════╡

# │ 0   ┆ 45.0 │

# ├╌╌╌╌╌┼╌╌╌╌╌╌┤

# │ 1   ┆ 63.0 │

# └─────┴──────┘

```