Skip to content

Commit

Permalink
start on github pages docs
Browse files Browse the repository at this point in the history
  • Loading branch information
nickzoic committed Oct 25, 2023
1 parent 1a7dac1 commit 5dbd335
Show file tree
Hide file tree
Showing 4 changed files with 123 additions and 3 deletions.
Empty file.
19 changes: 16 additions & 3 deletions docs/index.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,20 @@
# CountESS Documentation

[Github Repository](https://github.com/CountESS-Project/CountESS/)
This is CountESS, a modular, Python 3 reimplementation of Enrich2.

Documentation:
[Source code is available](https://github.com/CountESS-Project/CountESS)
and contributions are welcomed.

[This documentation is generated](https://github.com/CountESS-Project/CountESS/tree/main/docs)
from the `/docs/` directory and documentation fixes are also welcomed!

## Documentation

* [Installing CountESS](installing-countess/)
* [Writing Plugins](writing-plugins/)
* [Configuration File Format](config-file-format/)

## Presentations

* [CountESS: Count-based Experimental Scoring and Statistics](https://www.youtube.com/watch?v=JzU6cbvZ0a0) at [PyConAU 2023](https://2023.pycon.org.au/)

* [plugins](plugins/)
25 changes: 25 additions & 0 deletions docs/installing-countess/index.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,25 @@
## Installing CountESS

CountESS can be installed from pypi:
```
pip install CountESS
```

... or install the latest development version directly from github:
```
pip install git+https://github.com/CountESS-Project/CountESS.git
```

... or download and install for development:
```
git clone https://github.com/CountESS-Project/CountESS.git
cd CountESS
pip install -e .
```

... or run with nix:
```
nix run github:CountESS-Project/CountESS
```


82 changes: 82 additions & 0 deletions docs/writing-plugins/index.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,82 @@
# Writing CountESS Plugins

CountESS is a framework which connects together many components which do the actual
work, these components are called Plugins.

There are many plugins supplied as part of CountESS but it is also very easy to
write your own.

## General Structure

Plugins are provided by Python classes.
All CountESS plugin classes are required to be subclasses of
`countess.core.plugins.BasePlugin`.

CountESS detects installed plugins using the
[Python `importlib.metadata.entry_points`](https://docs.python.org/3/library/importlib.metadata.html#entry-points)
mechanism. So to declare your new class as a CountESS plugin, you
just need to add a clause like this to your `pyproject.toml` or
equivalent:

```
[project.entry-points.countess_plugins]
myplugin = "myproject:MyPluginClass"
```

## Types of Plugin

The most common kind of plugin takes a pandas dataframe, transforms it,
and passes the transformed dataframe along to the next plugin.

Examples include the Expression plugin, which creates a new column by
applying a formula to the existing columns, or the Regex Tool plugin which
uses a regular expression to break an existing column into multiple
parts.

Other types of plugin include readers, which read from a file in
some particular format, and writers, which write formatted data.

## Configuration Parameters

Plugins have "configuration parameters" which are declared in the plugin
class declaration using subclasses of `countess.core.parameters.BaseParameter`.

Configfuration values are provided by a [single configuration file which
is read and written by CountESS](../config-file-format/)

## Subclassing `BasePlugin`

To make life easier, there are several subclasses of `BasePlugin`
under `countess.core.plugins` ... each of these provides common
features used in different types of plugins.

### PandasSimplePlugin

### PandasTransform**X**To**Y**Plugin

There are several classes of the form `PandasTransform`**X**`To`**Y**`Plugin`.

The most obviously useful ones are `PandasTransformSingleToSinglePlugin`,
which translates a single value into another single value, and
`PandasTransformDictToDictPlugin`, which translates an entire table row.

To use these classes, subclass the appropriate one for your plugin and
override `process_value`, `process_row` or `process_dict` as appropriate.


## Built-in Plugins

### `LoadCSVPlugin`

### `LoadHdfPlugin`

### `LoadFastqPlugin`

### `LogScorePlugin`

### `GroupByPlugin`

## Example Plugins

Under [example-plugins](../example-plugins/) are files showing the
suggested layout for Python code in a plugin.

0 comments on commit 5dbd335

Please sign in to comment.