Update app.py

app.py

@@ -80,38 +80,40 @@ with colT2:
     font-size:16px ; font-family: 'Times New Roman'; color: #3358ff;} 
     </style> """, unsafe_allow_html=True)
 
-def analizar_tweets(search_words, number_of_tweets ):
-  tweets = api.user_timeline(screen_name = search_words,tweet_mode="extended", count= number_of_tweets)
-  result = []
-  for tweet in tweets:
-      if (tweet.full_text.startswith('RT')):
-          continue
-      elif not tweet.full_text.strip():
-          continue
-      else:
-          datos = preprocess(tweet.full_text)
-          prediction = pipeline_nlp(datos)
-          for predic in prediction:
-              etiqueta = {'Tweets': datos,'Prediccion': predic['label'], 'Probabilidad': predic['score']}
-              result.append(etiqueta)
-              
-  df = pd.DataFrame(result)
-  df['Prediccion'] = np.where( df['Prediccion'] == 'LABEL_1', 'Sexista', 'No Sexista')
-  df = df[df["Prediccion"] == 'Sexista']
-  df = df[df["Probabilidad"] > 0.80]
-  tabla = st.table(df.reset_index(drop=True).head(30).style.applymap(color_survived, subset=['Prediccion']))
-  resultado=df.groupby('Prediccion')['Probabilidad'].sum()
-  colores=["#aae977","#EE3555"]
-  fig, ax = plt.subplots(figsize=(2, 2), subplotpars=None)
-  plt.pie(resultado,labels=resultado.index,autopct='%1.1f%%',colors=colores)
-  ax.set_title("Porcentajes por Categorias", fontsize=8, fontweight="bold")
-  plt.rcParams.update({'font.size':8, 'font.weight':'bold'})
-  ax.legend()
-  # Muestra el gráfico
-  plt.show()
-  st.set_option('deprecation.showPyplotGlobalUse', False)
-  st.pyplot()
-    
+def analizar_tweets(search_words, number_of_tweets):
+  if(number_of_tweets > 0):
+      tweets = api.user_timeline(screen_name = search_words,tweet_mode="extended", count= number_of_tweets)
+      result = []
+      for tweet in tweets:
+          if (tweet.full_text.startswith('RT')):
+              continue
+          elif not tweet.full_text.strip():
+              continue
+          else:
+              datos = preprocess(tweet.full_text)
+              prediction = pipeline_nlp(datos)
+              for predic in prediction:
+                  etiqueta = {'Tweets': datos,'Prediccion': predic['label'], 'Probabilidad': predic['score']}
+                  result.append(etiqueta)            
+      df = pd.DataFrame(result)
+      df['Prediccion'] = np.where( df['Prediccion'] == 'LABEL_1', 'Sexista', 'No Sexista')
+      df = df[df["Prediccion"] == 'Sexista']
+      df = df[df["Probabilidad"] > 0.5]
+      tabla = st.table(df.reset_index(drop=True).head(30).style.applymap(color_survived, subset=['Prediccion']))
+      resultado=df.groupby('Prediccion')['Probabilidad'].sum()
+      colores=["#aae977","#EE3555"]
+      fig, ax = plt.subplots(figsize=(1, 1), subplotpars=None)
+      plt.pie(resultado,labels=resultado.index,autopct='%1.1f%%',colors=colores)
+      ax.set_title("Porcentajes por Categorias", fontsize=1, fontweight="bold")
+      plt.rcParams.update({'font.size':1, 'font.weight':'bold'})
+      ax.legend()
+      # Muestra el gráfico
+      plt.show()
+      st.set_option('deprecation.showPyplotGlobalUse', False)
+      st.pyplot()
+  else:
+      tabla = st.text("Ingrese los valores")
+      
   return tabla
 
 def tweets_localidad(buscar_localidad):