Upload scattertext_funtion.py

scattertext_funtion.py

@@ -0,0 +1,63 @@
+
+import streamlit as st
+import scattertext as stx
+import spacy
+import pandas as pd
+import en_core_web_sm
+
+
+
+# load language model:
+nlp = en_core_web_sm.load()
+nlp = spacy.load("en_core_web_md")
+
+# config
+st.title("Scattertext Analysis")
+# TODO：update other web settings
+
+# upload file
+uploaded_file = st.file_uploader("Upload your text document", type=["csv", "txt"])
+# proceed data
+if uploaded_file is not None:
+    if uploaded_file.name.endswith(".csv"):
+        df = pd.read_csv(uploaded_file)
+    elif uploaded_file.name.endswith(".txt"):
+        df = pd.read_csv(uploaded_file, sep='\t')
+        # TODO: check, assume contents are seperated by Tabs.
+# choose function
+    function_choice = st.selectbox('Choose function', ('Choose...', 'Generate Scattertext Plot', '...'))
+    # TODO: add new functions here (paper sections)
+# define function1
+    if function_choice == 'Generate Scattertext Plot':
+        text_columns = df.select_dtypes(include=['object']).columns.tolist()
+        chosen_column = st.selectbox("Choose text column for analysis", text_columns)
+        if st.button('Generate Scattertext Plot'):
+            # convert to scattertext corpus
+            corpus = stx.CorpusFromPandas(
+                text_columns,
+                category_col=text_columns[chosen_column].head(),
+                text_col=chosen_column,
+                nlp=nlp,
+                ). build()
+            # create visualization
+            # customize parameters
+            input_category_name = input('Enter the category name')
+            customize_category_name = st.text_input('Customize parameter', input_category_name)
+            input_non_category_name = input('Enter the non-category name')
+            customize_non_category_name = st.text_input('Customize parameter', input_non_category_name)
+            html = stx.produce_scattertext_explorer(corpus,
+                                       category=df[chosen_column].head(),
+                                       category_name=customize_category_name,
+                                       not_category_name=customize_non_category_name,
+                                       width_in_pixels=1000,
+                                       minimum_term_frequency=0,
+                                       metadata=df)
+        st.components.v1.html(html)
+# TODO: insert new functions
+
+    else:
+        st.error("Unsupported file format.")
+else:
+    st.write("Please upload a CSV or TXT file to begin.")
+
+