Skip to main content

A Scheme kernel for Jupyter that can use Python libraries

Project description

# Calysto Scheme

You can try Calysto Scheme without installing anything by clicking on the following button:

[![Binder](https://mybinder.org/badge.svg)](https://mybinder.org/v2/gh/Calysto/calysto_scheme/master?filepath=notebooks%2FReference%20Guide%20for%20Calysto%20Scheme.ipynb)

Calysto Scheme is a real Scheme programming language, with full support for continuations, including call/cc. It can also use all Python libraries. Also has some extensions that make it more useful (stepper-debugger, choose/fail, stack traces), or make it better integrated with Python. For more details on using Calysto Scheme, see:

http://nbviewer.jupyter.org/github/Calysto/calysto_scheme/blob/master/notebooks/Reference%20Guide%20for%20Calysto%20Scheme.ipynb

In Jupyter notebooks, because Calysto Scheme uses [MetaKernel](https://github.com/Calysto/metakernel/blob/master/README.rst), it has a fully-supported set of “magics”—meta-commands for additional functionality. This includes running Scheme in parallel. See all of the [MetaKernel Magics](https://github.com/Calysto/metakernel/blob/master/metakernel/magics/README.md).

Calysto Scheme is written in Scheme, and then translated into Python (and other backends). The entire functionality lies in a single Python file: https://github.com/Calysto/calysto_scheme/blob/master/calysto_scheme/scheme.py However, you can easily install it (see below).

Calysto Scheme in use:

## Parallel Processing

To use Calysto Scheme in parallel, do the following:

  1. Make sure that the Python module ipyparallel is installed. In the shell, type:

` pip install ipyparallel `

  1. To enable the extension in the notebook, in the shell, type:

` ipcluster nbextension enable `

  1. To start up a cluster, with 10 nodes, on a local IP address, in the shell, type:

` ipcluster start --n=10 --ip=192.168.1.108 `

  1. Initialize the code to use the 10 nodes, inside the notebook from a host kernel (can be any metakernel kernel), type:

` %parallel calysto_scheme CalystoScheme `

  1. Run code in parallel, inside the notebook, type:

Execute a single line, in parallel:

` %px (+ 1 1) `

Or execute the entire cell, in parallel:

` %%px (* cluster_rank cluster_rank) `

Results come back in a Scheme vector, in cluster_rank order. Therefore, the above would produce the result:

`scheme #10(0 1 4 9 16 25 36 49 64 81) ` You can get the results back in the host Scheme by accessing the variable _ (single underscore).

Notice that you can use the variable cluster_rank to partition parts of a problem so that each node is working on something different.

In the examples above, use -e to evaluate the code in the host Scheme as well. Note that cluster_rank is not defined on the host machine, and that this assumes the host kernel is the same as the parallel machines.

A full notebook example can be found here: [Mandelbrot.ipynb](https://github.com/Calysto/metakernel/blob/master/examples/Mandelbrot.ipynb)

## Install

Ensure that your system has jupyter installed, and pyQt and qtconsole if you want to use the qtconsole option.

You can install Calysto Scheme with Python3:

` pip3 install --upgrade calysto-scheme --user python3 -m calysto_scheme install --user `

or in the system kernel folder with:

` sudo pip3 install --upgrade calysto-scheme sudo python3 -m calysto_scheme install `

You can also use the –sys-prefix to install into your virtualenv.

Change pip3/python3 to use a different pip or Python. The version of Python used will determine how Calysto Scheme is run.

Use it in the Jupyter console, qtconsole, or notebook:

` jupyter console --kernel calysto_scheme jupyter qtconsole --kernel calysto_scheme jupyter notebook `

You can also just use the Python program, but it doesn’t have a fancy Read-Eval-Print Loop. Just run:

` python calysto_scheme/scheme.py `

## Requires

  • Python3

  • metakernel (installed automatically)

Calysto Scheme can also be un under PyPy for increased performance.

## Features

Calysto Scheme supports:

  • continuations

  • use of all Python libraries

  • choose/fail - built in fail and try again

  • produces stack trace (with line numbers), like Python

  • test suite

Planned:

  • Object-oriented class definitions and instance creation

  • complete Scheme functions (one can fall back to Python for now)

Limitations:

  • Runs slow on CPython; try PyPy

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distribution

calysto_scheme-1.4.8-py2.py3-none-any.whl (313.2 kB view details)

Uploaded Python 2 Python 3

File details

Details for the file calysto_scheme-1.4.8-py2.py3-none-any.whl.

File metadata

File hashes

Hashes for calysto_scheme-1.4.8-py2.py3-none-any.whl
Algorithm Hash digest
SHA256 f8c2c21fc9ae2ec28c766656b1647a22fc4f2a97d77606bc986712b8978fcb8d
MD5 fc75e71d99a7bcc7a7568dc78011553f
BLAKE2b-256 e018637cefec056ed337b11ed249f75d978a53a3db0660b7f5e4243f2bf70093

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page