Reference Guide

Extensions can be managed directly from within the pd.ext namespace. This page describes the functions available.

Repository Management

class pandex.ext.ExtensionRepository

This class allows the user to install and manage extensions that are be accessible from data frames.

Assuming that pandex is imported as pd, then the methods in this class are accessed from the namespace:

pd.ext.<method_name>

where <method_name> is one of:

disable_extension(name)

Disable an extension

Input Parameters:

name:

string name of extension (including collection if there is one)

enable_extension(name)

Enable an extension

Input Parameters:

name:

string name of extension (including collection if there is one)

import_extension(extension_spec)

Imports extensions that don’t already exist into the default repo

Input Parameters:

extension_spec:

A string with one or more lines of format:

extension_location -> name

where:

extension_location:

location of all the files needed for that extension. There must be a function with the extension name in one of the .py files at that location

Format options:

• <directory name on local file system>

• <.py file  name on local file system>

• Files on GitHub are referenced:

  github:username/repo[@branch/tag][/path/to/directory]
  

  where repo is a public repo on github for username. The latest commit on master is the default unless a specific branch or tag is optionally provided.

name:

string name of extension (including collection if required) to be installed

reinstall_extension(name)

Re-install an extension from the original source

Input Parameters:

name:

string name of extension (including collection if there is one)

remove_extension(name)

Permanently remove an extension from the repository

Input Parameters:

name:

string name of extension (including collection if there is one)

import statement

An alternative way is provided to install and manage multiple extensions in one go. This is useful, for example, to minimise the number of downloads when extensions are hosted on remote sites like GitHub.

class pandex.ext.ExtensionImporter

Extensions may also be to be defined in a text file and imported using the standard python import statement.

The text file must be have a file extension of .pdext and be located in the same directory as the file where the import is called from.

Format of text file:

# Can contain comment lines beginning with #

# Blank lines are ignored

# one extension specification per line
/path/to/extension/files  -> extension_function1
/path/to/extension/files  -> collection.extension_function2
github:username/repo      -> extension_function3

As an example, if a file named my_extensions .pdext is created with the extensions above, then a file named analysis.py can begin:

import pandex as pd
import my_extensions

df = pd.DataFrame({'x': [1,2,3,4]})

# call the extension
df.ext.extension_function1()