def interactive():
"""Page to allow nlp analysis from user input"""
input_text = st.text_area("Enter text", "Type here")
token_cb = st.checkbox("Show tokens")
ner_cb = st.checkbox("Show named entities")
sentiment_cb = st.checkbox("Show sentiment")
summary_cb = st.checkbox("Show Summary")
# st.success("Running Analysis")
# if st.button("Analysis"):
if token_cb:
tokens = az.tokenize(input_text)
st.write(tokens)
if ner_cb:
doc = az.get_nlp(input_text)
named_entities = az.named_entity_recognization(input_text)
if len(named_entities) > 0:
html = spacy.displacy.render(doc, style="ent")
# Newlines seem to mess with the rendering
html = html.replace("\n", " ")
HTML_WRAPPER = """<div style="overflow-x: auto; border: 1px solid \
#e6e9ef; border-radius: 0.25rem; padding: 1rem; margin-bottom: 2.5rem">\
{}</div>"""
st.write(HTML_WRAPPER.format(html), unsafe_allow_html=True)
else:
st.info("No named entity recognized")
if sentiment_cb:
sentiments = TextBlob(az.lemmatized_text(input_text))
st.write(sentiments.sentiment)
if summary_cb:
summaries = sz.summarize_text(input_text)
st.write(summaries)
def test_tokenize():
"""Test tokenize break down str into list of str correctly with the porter
method from nltk package"""
input_text = "Test tokenize break down str into list of str correctly"
output = az.tokenize(input_text)
expected = ["test", "tokenize", "break", "str", "list", "str", "correctly"]
assert output == expected
def df_preprocess(df):
"""build and preprocess (combine, normalize, tokenize) text"""
# filter out first two columns -- non-report content
cols = df.columns[2:]
# combining text into combined column
df["combined"] = df[cols].apply(
lambda row: "\n".join(row.values.astype(str)), axis=1)
# normalize
df[cts.NORMAL] = df["combined"].apply(lambda row: az.normalize(row))
# tokenize
df[cts.TOKEN] = df[cts.NORMAL].apply(lambda row: az.tokenize(row))
return df
def df_preprocess(df):
"""Build and preprocess (combine, normalize, tokenize) text."""
# filter out first two columns -- non-report content
# (student and assignment name)
cols = df.columns[2:]
# combining text into combined column
df[cts.COMBINED] = df[cols].apply(
lambda row: "\n".join(row.values.astype(str)), axis=1)
# normalize
df[cts.NORMAL] = df[cts.COMBINED].apply(lambda row: az.normalize(row))
# tokenize
df[cts.TOKEN] = df[cts.NORMAL].apply(lambda row: az.tokenize(row))
def interactive():
"""Page to allow nlp analysis from user input."""
input_text = st.text_area("Enter text", "Type here")
token_cb = st.checkbox("Show tokens")
ner_cb = st.checkbox("Show named entities")
sentiment_cb = st.checkbox("Show sentiment")
summary_cb = st.checkbox("Show Summary")
# st.success("Running Analysis")
# if st.button("Analysis"):
if token_cb:
tokens = az.tokenize(input_text)
st.write(tokens)
if ner_cb:
displacy_renderer(az.get_nlp(input_text))
if sentiment_cb:
sentiments = TextBlob(az.lemmatized_text(input_text))
st.write(sentiments.sentiment)
if summary_cb:
summaries = sz.summarize_text(input_text)
st.write(summaries)
def test_tokenize_parametrize(input_text, expected):
"""parametrize test tokenize"""
output = az.tokenize(input_text)
assert output == expected
def test_tokenize():
"""Test tokenize break down str into list of str correctly."""
input_text = "Test tokenize break down str into list of str correctly"
output = az.tokenize(input_text)
expected = ["test", "tokenize", "break", "str", "list", "str", "correctly"]
assert output == expected