Are you happy with your logging solution? Would you help us out by taking a 30-second survey? Click here


BayesDB on SQLite. A Bayesian database table for querying the probable implications of data as easily as SQL databases query the data itself.

Subscribe to updates I use bayeslite

Statistics on bayeslite

Number of watchers on Github 706
Number of open issues 194
Average time to close an issue 20 days
Main language Python
Average time to merge a PR 3 days
Open pull requests 9+
Closed pull requests 20+
Last commit over 1 year ago
Repo Created almost 5 years ago
Repo Last Updated over 1 year ago
Size 7.42 MB
Homepage http://probcomp.c...
Organization / Authorprobcomp
Page Updated
Do you use bayeslite? Leave a review!
View open issues (194)
View bayeslite activity
View on github
Fresh, new opensource launches 🚀🚀🚀
Trendy new open source projects in your inbox! View examples

Subscribe to our mailing list

Evaluating bayeslite for your project? Score Explanation
Commits Score (?)
Issues & PR Score (?)


Build Status Anaconda-Server Version Badge Anaconda-Server Installer Badge Anaconda-Server Platform Badge

BQL interpretation and storage for BayesDB. Please see for more information.


The easiest way to install bayeslite is to use the package on Anaconda Cloud. In your conda environment (python 2.7), run one of the following two commands:

$ conda install -c probcomp bayeslite             # latest release
$ conda install -c probcomp/label/edge bayeslite  # tip of master


Users and contributors should expect rapidly and dramatically shifting code and behavior at this time.



Our compatibility aim is to work on probcomp machines and members' laptops, and to provide scripts and instructions that make it not too hard to re-create our environments elsewhere. Pulls for polished packaging, broad installability, etc. are not appropriate contributions at this time.

Please run local tests before sending a pull request:

$ ./

That does not run the complete test suite, only the smoke tests, but is usually good enough. For the full suite:

$ ./ tests shell/tests


To build the documentation (requires sphinx):

$ make doc

The result will be placed in build/doc, with one subdirectory per output format.

To build only one output format, e.g. HTML because you don't want to install TeX:

$ make html
bayeslite open issues Ask a question     (View All Issues)
  • almost 3 years Is the Jenkins server forced to not skip any tests due to missing requirements?
  • almost 3 years Is the Jenkins server running the full test suite ie including __ci_ tests?
  • almost 3 years Eliminate copypasta of ast.Simulate and ast.CreateTabSim
  • almost 3 years Fix bayeslite-pull-requests project on Jenkins CI Server to automatically retrieve remote branches
  • almost 3 years Implement INSERT INTO for metamodel
  • almost 3 years Crash on CREATE POPULATION with a column of all `nan` values causes issue(s)
  • almost 3 years Docs say CORRELATION returns Pearson coeff for numeric columns, but returns R^2
  • almost 3 years bdb.execute(sql) fails but bdb.sql_execute(sql) succeeds on SELECT
  • almost 3 years Improve handling of reasons in GUESS SCHEMA
  • almost 3 years Perform error checking when using the OPTIMIZED flag for MML ANALYZE
  • about 3 years SIMULATE with a latent variable fails when inside a CREATE TABLE clause
  • about 3 years Quick search box in BQL doc returns nothing for important keywords
  • about 3 years Elevate "generators" and "populations" to the same level in BQL
  • about 3 years Patch outdated/incorrect bayeslite terminology
  • about 3 years SQLError - too many SQL variables on table with 1000 columns
  • about 3 years gracefully report error in CGPM for latent categoricals without specified number of categories
  • about 3 years draft a scheme for deciding agreement of statistical type with distributions in some metamodel
  • about 3 years rename num_predictions to nsamples in parameters of IBayesDBMetamodel.simulate_joint
  • about 3 years Implement BQL syntax for multiple/joint imputation
  • about 3 years write fast automatic tests for all new language features of 0.2
  • about 3 years add CGPM tests that do not require sklearn/venture/&c.
  • about 3 years mark CGPM tests as needing sklearn/venture/&c., and make them optional
  • about 3 years pass query parameters through to metamodel syntax
  • about 3 years latent variables cannot be inputs to CGPM foreign models
  • about 3 years eliminate core.bayesdb_generator_column_*
  • about 3 years find all the XXX's and convert them into issues
  • about 3 years omit needless table bayesdb_generator_column
  • about 3 years Sort out confusing mappings between categorical variables
  • about 3 years Support CGPMs when there is no default metamodel
  • about 3 years use virtual-table-valued functions for SIMULATE
bayeslite open pull requests (View All Pulls)
  • Transmit BayesDB stochasticity to crosscat engine
  • FOR COMMENT ONLY: persistence layer proposal
  • Distinguish 'single'- and "double"-quoted tokens in CREATE GENERATOR.
  • Fix a couple of typos.
  • WIP: fixes for pip install -e
  • Adding more stat types that are supported by CGPM
  • 20180301 schaechtle cmi with loom
  • TEST: do not merge
  • [WIP] Inference quality tests
bayeslite list of languages used
More projects by probcomp View all
Other projects in Python