Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/chyikwei/bnp
Bayesian nonparametric models for python
https://github.com/chyikwei/bnp
bayesian data-analysis probabilistic-graphical-models python topic-modeling
Last synced: about 2 months ago
JSON representation
Bayesian nonparametric models for python
- Host: GitHub
- URL: https://github.com/chyikwei/bnp
- Owner: chyikwei
- Created: 2017-03-04T04:08:08.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2018-09-11T14:39:53.000Z (about 6 years ago)
- Last Synced: 2024-04-08T02:57:17.709Z (6 months ago)
- Topics: bayesian, data-analysis, probabilistic-graphical-models, python, topic-modeling
- Language: Python
- Homepage:
- Size: 6.44 MB
- Stars: 17
- Watchers: 2
- Forks: 4
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome-topic-models - bnp - Cython reimplementation based on *online-hdp* following scikit-learn's API. (Models / Hierarchical Dirichlet Process (HDP) [:page_facing_up:](https://papers.nips.cc/paper/2004/file/fb4ab556bc42d6f0ee0f9e24ec4d1af0-Paper.pdf))
README
[](https://travis-ci.org/chyikwei/bnp)
[](https://circleci.com/gh/gh/chyikwei/bnp)
[](https://coveralls.io/github/chyikwei/bnp?branch=master)
# Bayesian Nonparametric
Bayesian Nonparametric models with Python.
Models follow scikit-learn's API and can be used as its extension.
Current model:
--------------
- **Hierarchical Dirichlet Process**
HDP is similar to LDA (Latent Direchlet Allocation) but assumes an "infinite" number of topics. This implementation is based on Chong Wang's online-hdp and optimized with cython.
Reference:
----------
- "Stochastic Variational Inference", Matthew D. Hoffman, David M. Blei, Chong Wang, John Paisley, 2013
- "Online Variational Inference for the Hierarchical Dirichlet Process", Chong Wang, John Paisley, David M. Blei, 2011
- Chong Wang's [online-hdp code](https://github.com/blei-lab/online-hdp).
Install:
--------
```
# clone repoisitory
git clone [email protected]:chyikwei/bnp.git
cd bnp
# install dependencies (cython, numpy, scipy, scikit-learn)
pip install -r requirements.txt
pip install .
```
Getting started:
----------------
In `bnp.utils` we proivde a function to generate fake document-word matrix with hidden topics. We will run our HDP model with it.
First, we can generate a document-word matrix with 5 hidden topics. (each topic has 10 uniuque words and each topic has 100 docs.)
```python
>>> from __future__ import print_function
>>> from bnp.online_hdp import HierarchicalDirichletProcess
>>> from bnp.utils import make_doc_word_matrix
>>> tf = make_doc_word_matrix(n_topics=5,
... words_per_topic=10,
... docs_per_topic=100,
... words_per_doc=20,
... shuffle=True,
... random_state=0)
>>> tf.shape
(500, 50)
```
For samples in the matrix, each row(document) only contains words from a specific topic (word 0 to 9: topic 1, 10 to 19: topic 2,...)
```python
>>> tf[0].toarray()
array([[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 2, 2, 1, 4, 1, 2, 3, 3, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0]])
>>> tf[1].toarray()
array([[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 3, 2, 3, 1, 3, 2, 1, 2, 0, 3, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0]])
```
Next we fit a HDP model with this matrix
```python
>>> hdp = HierarchicalDirichletProcess(n_topic_truncate=10,
... n_doc_truncate=3,
... max_iter=5,
... random_state=0)
>>> hdp.fit(tf)
```
Then we can print out topic proportion and top topic words in HDP model.
```python
# print topic function
>>> def print_top_words(model, n_words):
... topic_distr = model.topic_distribution()
... for topic_idx in range(model.lambda_.shape[0]):
... topic = model.lambda_[topic_idx, :]
... message = "Topic %d (proportion: %.2f): " % (topic_idx, topic_distr[topic_idx])
... message += " ".join([str(i) for i in topic.argsort()[:-n_words - 1:-1]])
... print(message)
>>> print_top_words(hdp, 10)
Topic 0 (proportion: 0.20): 3 1 7 5 8 4 0 2 9 6
Topic 1 (proportion: 0.00): 49 12 22 21 20 19 18 17 16 15
Topic 2 (proportion: 0.04): 43 49 44 45 47 40 46 48 41 42
Topic 3 (proportion: 0.13): 14 18 10 15 16 12 17 19 11 13
Topic 4 (proportion: 0.07): 19 16 10 15 11 17 12 13 18 14
Topic 5 (proportion: 0.01): 23 29 28 20 21 25 26 24 27 22
Topic 6 (proportion: 0.01): 31 38 35 39 30 33 34 37 32 36
Topic 7 (proportion: 0.19): 35 31 39 30 33 38 32 34 36 37
Topic 8 (proportion: 0.16): 48 42 46 49 45 47 41 44 40 43
Topic 9 (proportion: 0.19): 21 29 28 23 20 24 26 27 25 22
```
Here HDP find 7 large topics (> 1%) and those can map to the hidden topics we generated before.
Examples
--------
In `bnp/examples` folder. (Will add ipython notebook soon)
Running Test:
-------------
```
python setup.py test
```
Uninstall:
----------
```
pip uninstall bnp
```