pgmpy is a python library for working with Probabilistic Graphical Models.
Documentation and list of algorithms we support is at our official site http://pgmpy.org/
Our mailing list is at https://groups.google.com/forum/#!forum/pgmpy .
We have our community chat at gitter.
pgmpy has following non optional dependencies:
- Python 2.7 or Python 3
- NetworkX 1.9.1
- Scipy 0.12.1
- Numpy 1.9.2
- Cython 0.21
- Pandas 0.15.1
Currently pgmpy is not hosted on pypi or conda. You can either clone the git repo with:
git clone https://github.com/pgmpy/pgmpy
or download a zip from: https://github.com/pgmpy/pgmpy/archive/dev.zip
To install the dependencies switch to the pgmpy directory using:
$ cd /path/to/pgmpy
In the directory run either of the following:
Using pip
$ pip install -r requirements.txt # or requirements-dev.txt if you want to run unittests
or conda
$ conda install --file requirements.txt # or requirements-dev.txt
Then install using:
sudo python3 setup.py installIf you face any problems during installation let us know, via issues, mail or at our gitter channel.
You can check the latest sources from our github repository use the command:
git clone https://github.com/pgmpy/pgmpy.git
Issues can be reported at our issues section or via mail, or gitter. We will try our best to solve the issue at the earliest.
Before opening a pull request , have look at our contributing guide
Contributing guide contains some points that will make our life's easier in reviewing and merging your PR.
If you face any problems in pull request, feel free to ask them at mail or gitter.
If you have any new features, please have a discussion on the issue tracker or the mailing list before working on it.
After installation, you can launch the test form pgmpy
source directory (you will need to have the nose package installed):
$ nosetests -vto see the coverage of existing code use following command
$ nosetests --with-coverage --cover-package=pgmpy
Everything is at: http://pgmpy.org/
You can also build the documentation in your local system. We use sphinx to help us building documentation from our code.
$ cd /path/to/pgmpy/docs
$ make html
Then the docs will be in _build/html
Here is a small snippet of pgmpy API
python3
>>> from pgmpy.models import BayesianModel
>>> from pgmpy.factors import TabularCPD
>>> student = BayesianModel()
>>> # instantiates a new Bayesian Model called 'student'
>>> student.add_nodes_from(['diff', 'intel', 'grade'])
>>> # adds nodes labelled 'diff', 'intel', 'grade' to student
>>> student.add_edges_from([('diff', 'grade'), ('intel', 'grade')])
>>> # adds directed edges from 'diff' to 'grade' and 'intel' to 'grade'
>>> """
... diff cpd:
...
... +-------+--------+
... |diff: | |
... +-------+--------+
... |easy | 0.2 |
... +-------+--------+
... |hard | 0.8 |
... +-------+--------+
... """
>>> diff_cpd = TabularCPD('diff', 2, [[0.2], [0.8]])
>>> """
... intel cpd:
...
... +-------+--------+
... |intel: | |
... +-------+--------+
... |dumb | 0.5 |
... +-------+--------+
... |avg | 0.3 |
... +-------+--------+
... |smart | 0.2 |
... +-------+--------+
... """
>>> intel_cpd = TabularCPD('intel', 3, [[0.5], [0.3], [0.2]])
>>> """
... grade cpd:
...
... +------+-----------------------+---------------------+
... |diff: | easy | hard |
... +------+------+------+---------+------+------+-------+
... |intel:| dumb | avg | smart | dumb | avg | smart |
... +------+------+------+---------+------+------+-------+
... |gradeA| 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 |
... +------+------+------+---------+------+------+-------+
... |gradeB| 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 |
... +------+------+------+---------+------+------+-------+
... |gradeC| 0.8 | 0.8 | 0.8 | 0.8 | 0.8 | 0.8 |
... +------+------+------+---------+------+------+-------+
... """
>>> grade_cpd = TabularCPD('grade', 3,
[[0.1,0.1,0.1,0.1,0.1,0.1],
[0.1,0.1,0.1,0.1,0.1,0.1],
[0.8,0.8,0.8,0.8,0.8,0.8]],
evidence=['intel', 'diff'],
evidence_card=[3, 2])
>>> student.add_cpds(diff_cpd, intel_cpd, grade_cpd)
>>> # Finding active trail
>>> student.active_trail_nodes('diff')
{'diff', 'grade'}
>>> # Finding active trail with observation
>>> student.active_trail_nodes('diff', observed='grade')
{'diff', 'intel'}
pgmpy is released under MIT License. You can read about our lisence at here
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent