Update app.py

app.py

@@ -4,7 +4,7 @@ import gradio as gr
 import os
 import pandas as pd
 from datasets import load_dataset
-from sklearn.metrics.pairwise import cosine_similarity, pairwise_distances
+from sklearn.metrics.pairwise import cosine_similarity
 from datasets import Features, Value
 import plotly.express as px
 
@@ -16,7 +16,6 @@ Secret_token = os.getenv('HF_token')
 dataset = load_dataset("FDSRashid/embed_matn", token = Secret_token)
 books = load_dataset('FDSRashid/Hadith_info', data_files='Books.csv', token=Secret_token)['train'].to_pandas()
 df = dataset["train"].to_pandas()
-choice =  ['cityblock', 'cosine', 'euclidean', 'l1', 'l2', 'manhattan', 'canberra', 'chebyshev']
 
 dataset = load_dataset("FDSRashid/hadith_info", data_files = 'All_Matns.csv',token = Secret_token, features = features)
 matn_info = dataset['train'].to_pandas()
@@ -32,19 +31,17 @@ matn_info = pd.merge(matn_info, books, on='Book_ID')
 
 matn_info = matn_info.reset_index()
 df = df.reset_index()
-
 cols_to_use = df.columns.difference(matn_info.columns)
-
 joined_df = pd.merge(matn_info,df[cols_to_use],left_index=True, right_index=True)
 df = joined_df.copy()
 taraf_max = np.max(df['taraf_ID'].unique())
 
-def plot_similarity_score(taraf_num, metr):
+def plot_similarity_score(taraf_num):
     taraf_df = df[df['taraf_ID']== taraf_num]
     taraf_df['Number'] = np.arange(len(taraf_df))
     embed_taraf = taraf_df['embed'].to_list()
-    cos_score = pairwise_distances(embed_taraf, metric = metr)
-    fig = px.imshow(cos_score, color_continuous_scale='plasma_r')
+    cos_score = cosine_similarity(embed_taraf)
+    fig = px.imshow(cos_score)
     matr = cos_score
     rows, cols = matr.shape
     mask = np.tril(np.ones((rows, cols), dtype=bool), k=-1)
@@ -54,11 +51,8 @@ def plot_similarity_score(taraf_num, metr):
     return fig, fig_dis, taraf_df[['matn', 'Number', 'Book_Name', 'Author', 'Hadith Number']]
 
 with gr.Blocks() as demo: 
-  gr.Markdown('# Semantic Distance Visualizer')
-  gr.Markdown('Please note, the closer to zero, the better for this approach. This is semantic distance. On the Matrix plot, I\'ve reversed the color scales. This means that more yellow points are more similar, more purple = more dissimilar. ')
-    
+  gr.Markdown('# Semantic Similarity  Visualizer')
   taraf_number = gr.Slider(1,taraf_max , value=10000, label="Taraf", info="Choose the Taraf to Input", step = 1)
-  metric = gr.Dropdown(choices = choice, value = 'cosine', label = 'Variable to Display', info = 'Choose the variable to visualize.')
   btn = gr.Button('Submit')
-  btn.click(fn = plot_similarity_score, inputs = [taraf_number, metric], outputs = [gr.Plot(),gr.Plot(), gr.DataFrame()])
+  btn.click(fn = plot_similarity_score, inputs = [taraf_number], outputs = [gr.Plot(),gr.Plot(), gr.DataFrame()])
   demo.launch()