def empath(self, dataframe):
feat_builder = st.FeatsFromOnlyEmpath()
empath_corpus = st.CorpusFromParsedDocuments(
dataframe,
category_col='Document Type',
feats_from_spacy_doc=feat_builder,
parsed_col='Text').build()
html = st.produce_scattertext_explorer(
empath_corpus,
category='submission',
category_name='Submission',
not_category_name='Standard',
width_in_pixels=1000,
metadata=dataframe['Document'],
use_non_text_features=True,
use_full_doc=True,
topic_model_term_lists=feat_builder.get_top_model_term_lists())
logger.getLogger().info("Opening Empath Visual")
open(self.empath_file, 'wb').write(html.encode('utf-8'))
webbrowser.open("file://" + self.empath_file)
df_1 = df.groupby( [ "author"] ).size().reset_index(name='Counts')
df_1 = df_1.sort_values(by=['Counts'], ascending=False)
df_1 = df_1.head(2)
df_2 = df_1.merge(df, on='author', how='inner')
df_2 = df_2.sort_values(by=['Counts'], ascending=False)
build_corpus = st.CorpusFromPandas(df_2, category_col='author', text_col='clean_article_text', nlp=nlp).build()
df_freq = build_corpus.get_term_freq_df()
df_freq['GLOBE EDITORIAL SCORE'] = build_corpus.get_scaled_f_scores('GLOBE EDITORIAL')
df_freq['Jeffrey Simpson Score'] = build_corpus.get_scaled_f_scores('Jeffrey Simpson')
html = st.produce_scattertext_explorer(build_corpus,
category='GLOBE EDITORIAL',
category_name='GLOBE EDITORIAL',
not_category_name='Jeffrey Simpson',
width_in_pixels=1000,
metadata=df_2['author'])
open("../output/Top_2_Authors.html", 'wb').write(html.encode('utf-8'))
#visualizing Empath topics and categories instead of terms
build_feats = st.FeatsFromOnlyEmpath()
build_corpus_2 = st.CorpusFromParsedDocuments(df_2,category_col='author', feats_from_spacy_doc=build_feats, parsed_col='clean_article_text').build()
html = st.produce_scattertext_explorer(build_corpus_2,category='GLOBE EDITORIAL',category_name='GLOBE EDITORIAL',not_category_name='Jeffrey Simpson',width_in_pixels=1000,metadata=df_2['author'],use_non_text_features=True,use_full_doc=True,topic_model_term_lists=build_feats.get_top_model_term_lists())
open("../output/Top_2_Authors-Empath.html", 'wb').write(html.encode('utf-8'))
import scattertext as st
convention_df = st.SampleCorpora.ConventionData2012.get_data()
empath_feature_builder = st.FeatsFromOnlyEmpath()
corpus = st.CorpusFromPandas(
convention_df,
category_col='speaker',
text_col='text',
nlp=st.whitespace_nlp_with_sentences,
feats_from_spacy_doc=empath_feature_builder).build().get_unigram_corpus()
html = st.produce_pairplot(
corpus,
use_metadata=True,
category_projector=st.CategoryProjector(compactor=None),
topic_model_term_lists=empath_feature_builder.get_top_model_term_lists(),
metadata=convention_df['party'] + ': ' + convention_df['speaker'])
file_name = 'convention_pair_plot_empath.html'
open(file_name, 'wb').write(html.encode('utf-8'))
print('./' + file_name)
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import warnings
plt.style.use("seaborn")
warnings.filterwarnings(action='ignore')
listing = pd.read_csv('data/listings.csv')
listing['review_scores_ratings'] = listing['review_scores_rating'] / 20
listing['review_scores_ratings'] = pd.cut(
listing['review_scores_ratings'],
bins=5,
labels=["0.0-1.0", "1.0-2.0", "2.0-3.0", "3.0-4.0", "4.0-5.0"])
listing['description'].replace(np.NaN, "no_description", inplace=True)
listing["review_scores_ratings"].replace(np.NaN, "no_review", inplace=True)
nlp = spacy.load("en")
corpus = st.CorpusFromPandas(listing,
category_col="review_scores_rating",
text_col="description",
nlp=nlp).build().remove_terms(
ENGLISH_STOP_WORDS, ignore_absences=True)
# html=st.produce_sc
html = st.produce_scattertext_explorer(corpus,
category="4.0-5.0",
category_name="Star Rating 4.0 - 5.0",
not_category_name="All Other Ratings",
width_in_pixels=1000,
minimum_term_frequency=200)
open("term-associations.html", "wb").write(html.encode("utf-8"))
feat_builder = st.FeatsFromOnlyEmpath()