nlplot

Visualization Module for Natural Language Processing

Description

Facilitates the visualization of natural language processing and provides quicker analysis

You can draw the following graph

1. N-gram bar chart
2. N-gram tree Map
3. Histogram of the word count
4. wordcloud
5. co-occurrence networks
6. sunburst chart
7. pyLDAvis

（Tested in English and Japanese）

Requirement

• python package

Install

pip install nlplot

I've posted on this blog about the specific use. (Japanese)

And, The sample code is also available in the kernel of kaggle. (English)

Usage

sample df

df.head()

import nlplot

# target_col as a list type or a string separated by a space.
npt = nlplot.NLPlot(df, target_col='text')

# Stopword calculations can be performed.
stopwords = npt.get_stopword(top_n=30, min_freq=0)

# 1. N-gram bar chart
npt.bar_ngram(title='uni-gram', ngram=1, top_n=50, stopwords=stopwords)
npt.bar_ngram(title='bi-gram', ngram=2, top_n=50, stopwords=stopwords)

# 2. N-gram tree Map
npt.treemap(title='Tree of Most Common Words', ngram=1, top_n=30, stopwords=stopwords)

# 3. Histogram of the word count
npt.word_distribution(title='words distribution')

# 4. wordcloud
npt.wordcloud(stopwords=stopwords, colormap='tab20_r')

# 5. co-occurrence networks
npt.build_graph(stopwords=stopwords, min_edge_frequency=10)
# The number of nodes and edges to which this output is plotted.
# If this number is too large, plotting will take a long time, so adjust the [min_edge_frequency] well.
>> node_size:70, edge_size:166
npt.co_network(title='Co-occurrence network')

# 6. sunburst chart
npt.sunburst(title='sunburst chart', colorscale=True)

# 7. pyLDAvis
# If you want to run it in a notebook environment, you need to use the import and magic commands
import pyLDAvis
pyLDAvis.enable_notebook()
npt.ldavis(num_topics=5, passes=5, save=False)

Document

TBD

Test

cd tests
pytest

Other

• Plotly is used to plot the figure

  • https://plotly.com/python/

• co-occurrence networks is used to calculate the co-occurrence network

  • https://networkx.github.io/documentation/stable/tutorial.html

• The following is used to plot pyLDAvis

  • https://github.com/bmabey/pyLDAvis

• wordcloud uses the following fonts

  • https://mplus-fonts.osdn.jp/about.html