Added more features.

app.py

@@ -9,15 +9,19 @@ st.title('Sentiment Analysis for Price Trend Prediction')
 st.header(f'Data based on Platts News and Insights Data')
 st.subheader(f'{datetime.datetime.now()}')
 
-news_category = st.selectbox("Select Market Movers Category", ("Crude Oil", "Light Ends", "Middle Distillates", "Heavy Distillates"))
+# news_category = st.selectbox("Select Market Movers Category", ("Crude Oil", "Light Ends", "Middle Distillates", "Heavy Distillates"))
 
-latest_news = prepare_df(pd.read_csv('data/results_platts_09082024_clean.csv'), news_category)
-top_news = prepare_df(pd.read_csv('data/topresults_platts_09082024_clean.csv'), news_category)
+news_categories = st.multiselect("Select desired Market Movers categories", 
+                                 ["Macroeconomic & Geopolitics", "Crude Oil", "Light Ends", "Middle Distillates", "Heavy Distillates", "Other"],
+                                 ["Macroeconomic & Geopolitics", "Crude Oil"])
+
+latest_news = prepare_df(pd.read_csv('data/results_platts_09082024_clean.csv'), news_categories)
+top_news = prepare_df(pd.read_csv('data/topresults_platts_09082024_clean.csv'), news_categories)
 
 df_news = pd.concat([latest_news, top_news], ignore_index=True).drop_duplicates(['headline'])
 
 df_mean = pd.DataFrame({
-    'headline' : ['MEAN OF ALL NEWS'],
+    'headline' : ['MEAN OF SELECTED NEWS'],
     'negative_score' : [df_news['negative_score'].mean()],
     'neutral_score' : [df_news['neutral_score'].mean()],
     'positive_score' : [df_news['positive_score'].mean()],
@@ -30,4 +34,7 @@ df_news_final.index = np.arange(1, len(df_news_final) + 1)
 
 df_news_final
 
-st.plotly_chart(plot_3dgraph(df_news_final), use_container_width=True)
+try:
+    st.plotly_chart(plot_3dgraph(df_news_final), use_container_width=True)
+except:
+    st.subheader('Select news categories to plot 3D graph')

modules/data_preparation.py

@@ -2,6 +2,8 @@ import re
 import datetime
 import plotly.express as px
 import datetime
+import plotly.graph_objects as go
+import numpy as np
 
 def clean_text(text):
     new_text = text
@@ -9,29 +11,51 @@ def clean_text(text):
         new_text = re.sub(rgx_match, '', new_text)
     return new_text
 
-def prepare_df(df, category):
+def prepare_df(df, categories):
     try:
         df.drop(columns=['Unnamed: 0'], inplace=True)
     except:
         pass
 
-    if category == 'Crude Oil':
-        news_data = df[(df['topic_verification'] == 'Crude Oil') | (df['topic_verification'] == 'Macroeconomic & Geopolitics')]
-    if category == 'Light Ends':
-        news_data = df[(df['topic_verification'] == 'Light Ends')]
-    if category == 'Middle Distillates':
-        news_data = df[(df['topic_verification'] == 'Middle Distillates')]
-    if category == 'Heavy Distillates':
-        news_data = df[(df['topic_verification'] == 'Heavy Distillates')]
+    df['topic_verification'][(df.headline.str.contains('crude', case=False)) | df.body.str.contains('crude', case=False)] = 'Crude Oil'
 
-    actual_day = datetime.date.today() - datetime.timedelta(days=1)
-    pattern_del = actual_day.strftime('%b').upper()
+    try:
+        news_data = df[df['topic_verification'].isin(categories)]
+
+        actual_day = datetime.date.today() - datetime.timedelta(days=1)
+        pattern_del = actual_day.strftime('%b').upper()
 
-    filter = news_data['headline'].str.contains(pattern_del)
-    news_data = news_data[~filter]
+        filter = news_data['headline'].str.contains(pattern_del)
+        news_data = news_data[~filter]
+    
+    except Exception as E:
+        print(E)
 
     return news_data
 
+# def prepare_df(df, category):
+#     try:
+#         df.drop(columns=['Unnamed: 0'], inplace=True)
+#     except:
+#         pass
+
+#     if category == 'Crude Oil':
+#         news_data = df[(df['topic_verification'] == 'Crude Oil') | (df['topic_verification'] == 'Macroeconomic & Geopolitics')]
+#     if category == 'Light Ends':
+#         news_data = df[(df['topic_verification'] == 'Light Ends')]
+#     if category == 'Middle Distillates':
+#         news_data = df[(df['topic_verification'] == 'Middle Distillates')]
+#     if category == 'Heavy Distillates':
+#         news_data = df[(df['topic_verification'] == 'Heavy Distillates')]
+
+#     actual_day = datetime.date.today() - datetime.timedelta(days=1)
+#     pattern_del = actual_day.strftime('%b').upper()
+
+#     filter = news_data['headline'].str.contains(pattern_del)
+#     news_data = news_data[~filter]
+
+#     return news_data
+
 def plot_3dgraph(news_data):
     fig = px.scatter_3d(news_data, 
                         x='neutral_score', 
@@ -46,10 +70,22 @@ def plot_3dgraph(news_data):
                         hover_data='topic_verification')
 
     fig.update_layout(
-        height=800,
+        height=600,
         title=dict(text=f"Platts News Sentiments ({datetime.datetime.now().strftime('%d/%m/%y')})<br><sup>Hover cursor on a datapoint to show news title</sup>", font=dict(size=35), automargin=False)
     )
 
     fig.update_traces(textfont_size=8)
 
+    trace=dict(type='scatter3d',
+           x=news_data.iloc[[-1]]['neutral_score'],
+           y=news_data.iloc[[-1]]['negative_score'],
+           z=news_data.iloc[[-1]]['positive_score'],
+           mode='markers',
+           name= 'MEAN OF SELECTED NEWS',
+           marker=dict(color=[f'rgb({0}, {250}, {200})' for _ in range(25)],
+                       size=10)
+          )
+
+    fig.add_trace(trace)
+
     return fig