tanish78 commited on
Commit
db51a4a
·
verified ·
1 Parent(s): d4b76a8

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +1 -1
app.py CHANGED
@@ -104,7 +104,7 @@ def generate_bar_chart(df, num_clusters_to_display):
104
  top_clusters = df['Cluster'].value_counts().index[1:num_clusters_to_display+1]
105
  df_top_clusters = df[df['Cluster'].isin(top_clusters)]
106
 
107
- cluster_top_words = df_top_clusters.groupby('Cluster')['texts'].apply(lambda x: ' '.join(x)).reset_index()
108
  cluster_top_words['top_word'] = cluster_top_words['texts'].apply(lambda x: pd.Series(x.split()).value_counts().index[0] if len(x.split()) > 0 else '')
109
  cluster_sizes = df_top_clusters['Cluster'].value_counts().reset_index()
110
  cluster_sizes.columns = ['Cluster', 'Count']
 
104
  top_clusters = df['Cluster'].value_counts().index[1:num_clusters_to_display+1]
105
  df_top_clusters = df[df['Cluster'].isin(top_clusters)]
106
 
107
+ cluster_top_words = df_top_clusters.groupby('Cluster', observed=False)['texts'].apply(lambda x: ' '.join(x)).reset_index()
108
  cluster_top_words['top_word'] = cluster_top_words['texts'].apply(lambda x: pd.Series(x.split()).value_counts().index[0] if len(x.split()) > 0 else '')
109
  cluster_sizes = df_top_clusters['Cluster'].value_counts().reset_index()
110
  cluster_sizes.columns = ['Cluster', 'Count']