Update app.py

app.py

@@ -23,10 +23,14 @@ mpl.rcParams.update(mpl.rcParamsDefault)
 df = pd.read_parquet('virus_ds.parquet')
 virus = df['Organism_Name'].unique()
 virus = {v: v for v in virus}
-df_new = pd.read_parquet("virus.parquet")
+df_new = pd.read_parquet("virus.parquet", columns= ['organism_name'])
 df_new = df_new.groupby('organism_name').apply(lambda x: x.head(100) if len(x) > 10 else None).reset_index(drop=True)
 filter_species = df_new.organism_name.value_counts().reset_index()[df_new.organism_name.value_counts().reset_index()['count'] > 40 ]['organism_name'][1:].tolist()
+
+df_old = pd.read_parquet("virus.parquet", columns =['seq', 'organism_name'])
+MASTER_DF = df_old[df_old['organism_name'].isin(filter_species)].copy()
 del df_new
+del df_old
 virus_new = {v: v for v in filter_species}
 loss_typesss = pd.read_csv("training_data_5.csv")['loss_type'].unique().tolist()
 model_typesss = pd.read_csv("training_data_5.csv")['model_type'].unique().tolist()
@@ -89,11 +93,7 @@ with ui.navset_card_tab(id="tab"):
 
         @render.plot()
         def plot_distro():
-            df = pd.read_parquet("virus.parquet")
-            df = df.groupby('organism_name').apply(lambda x: x.head(100) if len(x) > 10 else None).reset_index(drop=True)
-            filter_species = df.organism_name.value_counts().reset_index()[df.organism_name.value_counts().reset_index()['count'] > 40 ]['organism_name'][1:].tolist()
-            
-            df = df[df["organism_name"].isin(input.virus_selector_1())]
+            df = MASTER_DF[MASTER_DF["organism_name"].isin(input.virus_selector_1())].copy()
             grouped = df.groupby("organism_name")["seq"].apply(list)
             return plot_distrobutions(grouped, grouped.index, input.basepair())