# pandas-dataclasses
[![Release](https://img.shields.io/pypi/v/pandas-dataclasses?label=Release&color=cornflowerblue&style=flat-square)](https://pypi.org/project/pandas-dataclasses/)
[![Python](https://img.shields.io/pypi/pyversions/pandas-dataclasses?label=Python&color=cornflowerblue&style=flat-square)](https://pypi.org/project/pandas-dataclasses/)
[![Downloads](https://img.shields.io/pypi/dm/pandas-dataclasses?label=Downloads&color=cornflowerblue&style=flat-square)](https://pepy.tech/project/pandas-dataclasses)
[![DOI](https://img.shields.io/badge/DOI-10.5281/zenodo.6127352-cornflowerblue?style=flat-square)](https://doi.org/10.5281/zenodo.6127352)
[![Tests](https://img.shields.io/github/workflow/status/astropenguin/pandas-dataclasses/Tests?label=Tests&style=flat-square)](https://github.com/astropenguin/pandas-dataclasses/actions)
pandas data creation made easy by dataclass
## Overview
pandas-dataclass makes it easy to create [pandas] data (DataFrame and Series) by specifying their data types, attributes, and names using the Python's dataclass:
<details>
<summary>Click to see all imports</summary>
```python
from dataclasses import dataclass
from pandas_dataclasses import AsFrame, Data, Index
```
</details>
```python
@dataclass
class Weather(AsFrame):
"""Weather information."""
year: Index[int]
month: Index[int]
temp: Data[float]
wind: Data[float]
df = Weather.new(
[2020, 2020, 2021, 2021, 2022],
[1, 7, 1, 7, 1],
[7.1, 24.3, 5.4, 25.9, 4.9],
[2.4, 3.1, 2.3, 2.4, 2.6],
)
```
where `df` will become a DataFrame object like:
```
temp wind
year month
2020 1 7.1 2.4
7 24.3 3.1
2021 1 5.4 2.3
7 25.9 2.4
2022 1 4.9 2.6
```
### Features
- Specifying data types and names of each element in pandas data
- Specifying metadata stored in pandas data attributes (attrs)
- Support for hierarchical index and columns
- Support for custom factory for data creation
- Support for full [dataclass] features
- Support for static type check by [mypy] and [Pyright] ([Pylance])
### Installation
```bash
pip install pandas-dataclasses
```
## How it works
pandas-dataclasses provides you the following features:
- Type hints for dataclass fields (`Attr`, `Column`, `Data`, `Index`) to specify the data type and name of each element in pandas data
- Mix-in classes for dataclasses (`As`, `AsFrame`, `AsSeries`) to create pandas data by a classmethod (`new`) that takes the same arguments as dataclass initialization
When you call `new`, it will first create a dataclass object and then create a Series or DataFrame object from the dataclass object according the type hints and values in it.
In the example above, `df = Weather.new(...)` is thus equivalent to:
<details>
<summary>Click to see all imports</summary>
```python
from pandas_dataclasses import asframe
```
</details>
```python
obj = Weather([2020, ...], [1, ...], [7.1, ...], [2.4, ...])
df = asframe(obj)
```
where `asframe` is a conversion function.
pandas-dataclasses does not touch the dataclass object creation itself; this allows you to fully customize your dataclass before conversion by the dataclass features (`field`, `__post_init__`, ...).
## Basic usage
### DataFrame creation
As shown in the example above, a dataclass that has the `AsFrame` (or `AsDataFrame` as an alias) mix-in will create DataFrame objects:
<details>
<summary>Click to see all imports</summary>
```python
from dataclasses import dataclass
from pandas_dataclasses import AsFrame, Data, Index
```
</details>
```python
@dataclass
class Weather(AsFrame):
"""Weather information."""
year: Index[int]
month: Index[int]
temp: Data[float]
wind: Data[float]
df = Weather.new(...)
```
where fields typed by `Index` are *index fields*, each value of which will become an index or a part of a hierarchical index of a DataFrame object.
Fields typed by `Data` are *data fields*, each value of which will become a data column of a DataFrame object.
Fields typed by other types are just ignored in the DataFrame creation.
Each data or index will be cast to the data type specified in a type hint like `Index[int]`.
Use `Any` or `None` (like `Index[Any]`) if you do not want type casting.
See also [data typing rules](#data-typing-rules) for more examples.
By default, a field name (i.e. an argument name) is used for the name of corresponding data or index.
See also [custom naming](#custom-naming) and [naming rules](#naming-rules) if you want customization.
### Series creation
A dataclass that has the `AsSeries` mix-in will create Series objects:
<details>
<summary>Click to see all imports</summary>
```python
from dataclasses import dataclass
from pandas_dataclasses import AsSeries, Data, Index
```
</details>
```python
@dataclass
class Weather(AsSeries):
"""Weather information."""
year: Index[int]
month: Index[int]
temp: Data[float]
ser = Weather.new(...)
```
Unlike `AsFrame`, the second and subsequent data fields are ignored in the Series creation even if they exist.
Other rules are the same as for the DataFrame creation.
## Advanced usage
### Metadata storing
Fields typed by `Attr` are *attribute fields*, each value of which will become an item of attributes of a DataFrame or a Series object:
<details>
<summary>Click to see all imports</summary>
```python
from dataclasses import dataclass
from pandas_dataclasses import AsFrame, Attr, Data, Index
```
</details>
```python
@dataclass
class Weather(AsFrame):
"""Weather information."""
year: Index[int]
month: Index[int]
temp: Data[float]
wind: Data[float]
loc: Attr[str] = "Tokyo"
lon: Attr[float] = 139.69167
lat: Attr[float] = 35.68944
df = Weather.new(...)
```
where `df.attrs` will become like:
```python
{"loc": "Tokyo", "lon": 139.69167, "lat": 35.68944}
```
### Custom naming
The name of attribute, data, or index can be explicitly specified by adding a hashable annotation to the corresponding type:
<details>
<summary>Click to see all imports</summary>
```python
from dataclasses import dataclass
from typing import Annotated as Ann
from pandas_dataclasses import AsFrame, Attr, Data, Index
```
</details>
```python
@dataclass
class Weather(AsFrame):
"""Weather information."""
year: Ann[Index[int], "Year"]
month: Ann[Index[int], "Month"]
temp: Ann[Data[float], "Temperature (deg C)"]
wind: Ann[Data[float], "Wind speed (m/s)"]
loc: Ann[Attr[str], "Location"] = "Tokyo"
lon: Ann[Attr[float], "Longitude (deg)"] = 139.69167
lat: Ann[Attr[float], "Latitude (deg)"] = 35.68944
df = Weather.new(...)
```
where `df` and `df.attrs` will become like:
```
Temperature (deg C) Wind speed (m/s)
Year Month
2020 1 7.1 2.4
7 24.3 3.1
2021 1 5.4 2.3
7 25.9 2.4
2022 1 4.9 2.6
```
```python
{"Location": "Tokyo", "Longitude (deg)": 139.69167, "Latitude (deg)": 35.68944}
```
If an annotation is a [format string], it will be formatted by a dataclass object before the data creation:
<details>
<summary>Click to see all imports</summary>
```python
from dataclasses import dataclass
from typing import Annotated as Ann
from pandas_dataclasses import AsFrame, Data, Index
```
</details>
```python
@dataclass
class Weather(AsFrame):
"""Weather information."""
year: Ann[Index[int], "Year"]
month: Ann[Index[int], "Month"]
temp: Ann[Data[float], "Temperature ({.temp_unit})"]
wind: Ann[Data[float], "Wind speed ({.wind_unit})"]
temp_unit: str = "deg C"
wind_unit: str = "m/s"
df = Weather.new(..., temp_unit="deg F", wind_unit="km/h")
```
where units of the temperature and the wind speed will be dynamically updated (see also [naming rules](#naming-rules)).
### Hierarchical columns
Adding tuple annotations to