Larch: a Python tool for choice modeling
Python C++ Objective-C JavaScript CSS Batchfile Other
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
bin
condabuild/larch
doc
img
py
scripts
sqlite
src
v4
.gitattributes
.gitignore
.readthedocs.yml rtfd Nov 7, 2016
LICENSE
README.rst
build_configuration.py
bumpversion.cfg
requirements.txt rtfd Oct 27, 2016
seedling-35.yml
seedling-36.yml
setup.py
setup_apsw.py
setup_common.py allow build larchsqlite in other places Feb 3, 2016
setup_sqlite.py
setup_test.py
taiga-35.yml
taiga-36.yml Add python-docx to reqs Mar 3, 2017
taiga-rtfd.yml
taiga.yml require docutils Feb 26, 2017
taiga_mac.yml
taiga_win.yml Version bump, compatible with Python3.6 Feb 26, 2017
update_git_version.py
windows_build.bat windows builder for L: Nov 18, 2016

README.rst

larch

https://readthedocs.org/projects/larch/badge/?version=latest&style=round

Larch: the logit architect

This is a tool for the estimation and application of logit-based discrete choice models. It is designed to integrate with NumPy and facilitate fast processing of linear models. If you want to estimate non-linear models, try Biogeme, which is more flexible in form and can be used for almost any model structure. If you don't know what the difference is, you probably want to start with linear models.

This project is very much under development. There are plenty of undocumented functions and features; use them at your own risk. Undocumented features may be non-functional, not rigorously tested, deprecated or removed without notice in a future version. If a function or method is documented, it is intended to be stable in future updates.

FAQ

Why is the Windows download so much larger than the Mac download?

The Windows wheel include the openblas library for linear algebra computations. The Mac version does not need an extra library because Mac OS X includes vector math libraries by default.

It is not working. Can you troubleshoot for me?

Are you using the 64 bit (amd64) version of Python? Larch is only compiled for 64 bit at present.

For some unknown reason, certain mathematical tools are not available on PyPI as wheels for Windows. You will need to download numpy, scipy, and pandas and install them manually.

You may also need to install the Microsoft Visual C++ 2015 <https://www.microsoft.com/en-us/download/details.aspx?id=48145> redistributable libraries. Future versions of Larch may include these for you.