Code to generate Picasso embeddings of any input matrix. Picasso maps the points of an input matrix to user-defined, n-dimensional shape coordinates, while minimizing reconstruction error using an autoencoder neural network structure. In the sample code Picasso is applied to single-cell gene expression counts.
Examples for running Picasso can be found in examplePicasso.ipynb. The notebook can be run in Google Colab by clicking on the 
An introduction to using Colab can be found here. Briefly, run each code cell by selecting the cell and executing Command/Ctrl+Enter. Code cells can be edited by simply clicking on the cell to start typing.
Elephant coordinates generated from Mayer et al. 2010.
You need Python 3.6 or later to run Picasso. You can have multiple Python versions (2.x and 3.x) installed on the same system without problems.
In Ubuntu, Mint and Debian you can install Python 3 like this:
$ sudo apt-get install python3 python3-pip
For other Linux flavors, macOS and Windows, packages are available at
Clone this repo:
$ git clone https://github.com/pachterlab/picasso.git
$ cd picasso
The necessary environment can be installed:
$ conda env create -f env/env3.7_LINUX.yml
$ conda activate env3.7
Or for MACOS:
$ conda env create -f env/env3.7_MACOS.yml
Import the module to use as in the examplePicasso.ipynb:
>>> from Picasso import Picasso