Upload New File

p_2/code/pdp.py

+
+import pandas as pd
+pd.set_option('display.max_columns', 500)
+
+v = pd.read_excel(r"data(7).xlsx")
+v_list=[]
+for index,name in (enumerate(v['variant'].value_counts().index.tolist())):
+    v_list.append(name)
+v_list.append('not_sequenced')
+month_data_encoded=pd.read_excel(r"month_data_encoded.xlsx")
+data_encoded=pd.read_excel(r"data_encoded.xlsx")
+import tensorflow as tf
+import tensorflow_probability as tfp
+import numpy as np
+import matplotlib.pyplot as plt
+import seaborn as sns
+import time
+from sklearn.model_selection import train_test_split
+model = tf.keras.models.load_model(r"convmodel")
+def accuracy_score(preds, labels):
+    #return np.mean(np.argmax(preds, axis=1) == np.argmax(labels, axis=1))
+    #predictednumbers=np.argmax(preds, axis=1)
+    #realnumbers=np.argmax(labels, axis=1)
+    diff=np.subtract(preds, labels)
+    return np.sqrt(sum(np.power(diff, 2))/len(diff))
+month_codes=[]
+for index,name in (enumerate(data_encoded['month'].value_counts().index.tolist())):
+    #for name in['Germany']:
+        month=name
+        data_encoded_2=data_encoded.loc[(data_encoded['month'] == month)]
+        month_codes.append([month,[data_encoded_2.iloc[0]['month_0'],data_encoded_2.iloc[0]['month_1'] ,data_encoded_2.iloc[0]['month_2'] ,data_encoded_2.iloc[0]['month_3']  ] ])
+
+        
+        
+from statistics import mean
+from Big_list import Big_list
+filtered = []
+
+for i in Big_list:
+    if mean(i[1]) >= 6.3:
+        filtered.append(i)
+
+
+vv=[[i[0],mean(i[1])] for i in list(filtered)]
+def Sort(sub_li):
+   
+    l = len(sub_li)
+     
+    for i in range(0, l):
+        for j in range(0, l-i-1):
+             
+            if (sub_li[j][1] > sub_li[j + 1][1]):
+                tempo = sub_li[j]
+                sub_li[j] = sub_li[j + 1]
+                sub_li[j + 1] = tempo
+     
+    return sub_li
+ff=Sort(vv)
+ff=[ff[i][0] for i in range(0,(len(ff)-2))]
+
+
+
+feature_list=ff
+xxx_list=[]
+cat_nr=100
+def valuation_formula(step):
+    return step
+for index,name in (enumerate(month_data_encoded['countriesAndTerritories'].value_counts().index.tolist())):
+    #for name in['Germany']:
+        land=name
+        print(land)
+        month_data_encoded_land=month_data_encoded.loc[(month_data_encoded['countriesAndTerritories'] == land)]
+        mde=month_data_encoded_land
+        mde=mde.drop(['dateRep', 'cases' , 'Rescd' , '7days_before_mean','7days_after_mean', 'week' ,'year' , 'index', 'month', 'countriesAndTerritories', 'va' , 'vaccin'], axis=1)
+        mde = mde.reindex(sorted(mde.columns), axis=1)
+        percent_list=[]
+        number_list=[-10000]
+        
+
+        
+        test_origin=mde
+        xx_list=[]
+        test=test_origin.copy(deep=True)
+        for step in [i[1] for i in month_codes]:
+            test['month_0'] = test.apply(lambda row: step[0], axis=1)
+            test['month_1'] = test.apply(lambda row: step[1], axis=1)
+            test['month_2'] = test.apply(lambda row: step[2], axis=1)
+            test['month_3'] = test.apply(lambda row: step[3], axis=1)
+            
+            X_test = test.drop(labels = ["rp_zeitraum"],axis = 1)
+            X_test = X_test.values.reshape(-1,1,111,1)
+            
+            n_mc_run = 100
+            #med_prob_thres = 0.35
+
+            y_pred_logits_list = [model(X_test) for _ in range(n_mc_run)]  # a list of predicted logits
+            y_pred_prob_all = np.concatenate([tf.nn.softmax(y, axis=-1)[:, :, np.newaxis] for y in y_pred_logits_list], axis=-1)
+            y_predicted_list=[np.argmax([np.mean(y_pred_prob_all[idx][i])  for i in range (0,len(y_pred_prob_all[idx]))],axis=-1) for idx in range (0,len(X_test)) ]
+            
+            mean_predicted=np.mean(y_predicted_list)
+            xx_list.append(mean_predicted)
+        xxx_list.append([land,'month',[[i[0] for i in month_codes],xx_list]])
+        print([land,'month',[[i[0] for i in month_codes],xx_list]])
+        print(xx_list)
+        for xx in feature_list:
+            print(xx)
+            xx_list=[]
+            test=test_origin.copy(deep=True)
+            for step in [0,1.0]:
+                test[xx] = test.apply(lambda row: valuation_formula(step), axis=1)
+                if xx in v_list:
+                    vv_list=v_list.copy()
+                    vv_list.remove(xx)
+                    for i in vv_list:
+                        test[i] = test[i].multiply(1-step)
+                
+                X_test = test.drop(labels = ["rp_zeitraum"],axis = 1)
+                X_test = X_test.values.reshape(-1,1,111,1)
+                
+                n_mc_run = 100
+                #med_prob_thres = 0.35
+
+                y_pred_logits_list = [model(X_test) for _ in range(n_mc_run)]  # a list of predicted logits
+                y_pred_prob_all = np.concatenate([tf.nn.softmax(y, axis=-1)[:, :, np.newaxis] for y in y_pred_logits_list], axis=-1)
+                y_predicted_list=[np.argmax([np.mean(y_pred_prob_all[idx][i])  for i in range (0,len(y_pred_prob_all[idx]))],axis=-1) for idx in range (0,len(X_test)) ]
+                
+                mean_predicted=np.mean(y_predicted_list)
+                xx_list.append(mean_predicted)
+            xxx_list.append([land,xx,xx_list])
+            print([land,xx,xx_list])
+               
+with open('xxx_list_r.py', 'w') as w:
+    w.write('xxx_list_r = %s' % xxx_list)       
\ No newline at end of file