redhog / jupyter-petrova

jupyter-petrova

Explore dataframes, arrays, scipy and matplotlib interactively w/o coding

Installation

To install use pip:

$ pip install jupyter_petrova

For a development installation (requires Node.js and Yarn version 1),

$ git clone https://github.com/redhog/jupyter-petrova.git
$ cd jupyter-petrova
$ pip install -e .
$ jupyter nbextension install --py --symlink --overwrite --sys-prefix jupyter_petrova
$ jupyter nbextension enable --py --sys-prefix jupyter_petrova

When actively developing your extension for JupyterLab, run the command:

$ jupyter labextension develop --overwrite jupyter_petrova

Then you need to rebuild the JS when you make a code change:

$ cd js
$ yarn run build

You then need to refresh the JupyterLab page when your javascript changes.

Usage

In a notebook cell, enter:

from jupyter_petrova import *
g = Graph()
g

Use the "Add" field to select python functions to add. A good starting point might be skimage.io.imread and anything under skimage.filters. Click and drag to move boxes on the board. Click on a function box to set its input parameters and view its output. To set a parameter to the output of another box, select the input field for that parameter, then shift-click the other box.

To pre-populate the graph,

test = Task("skimage.io.imread", fname="test.jpeg")
filtered = Task("skimage.filters.edges.sobel", image=test)
g = Graph(tasks = {"test": test, "filtered": filtered})
g

Individual tasks can be accessed using the dictionary g.tasks. Each task has a property value that contains the output value of that task. The function name is available in the task property name, and the parameters in params.