Zrobione zadanie 07.

rozwiazanie.py

@@ -0,0 +1,109 @@
+import numpy as np
+import pandas as pd
+
+from sklearn.linear_model import LinearRegression
+from sklearn import preprocessing
+from sklearn.metrics import mean_squared_error
+
+data_train = pd.read_csv("train/train.tsv", header=None, sep='\t')
+#print(data_train[[1,2,5]])
+data_train[3] = data_train[3].astype('category')
+data_train[4] = data_train[4].astype('category')
+brand_codest = dict(enumerate(data_train[3].cat.categories))
+brand_codes = {y:x for x,y in brand_codest.items()}
+#data_train[3].map(data_train_codes)
+data_train.replace({3: brand_codes}, inplace=True)
+#print(brand_codes)
+#print(data_train[3])
+fuel_codest = dict(enumerate(data_train[4].cat.categories))
+fuel_codes = {y:x for x,y in fuel_codest.items()}
+#data_train[3].map(data_train_codes)
+data_train.replace({4: fuel_codes}, inplace=True)
+#print(fuel_codes)
+#print(data_train[4])
+# #Normalizacja danych
+# flcols = data_train[[0, 1, 2]].columns
+# x = data_train[[0, 1, 2]].values
+# # min_max_scaler = preprocessing.MinMaxScaler()
+# max_abs_scaler = preprocessing.MaxAbsScaler()
+# # x_scaled = min_max_scaler.fit_transform(x)
+# x_scaled = max_abs_scaler.fit_transform(x)
+# normcols = pd.DataFrame(x_scaled, columns=flcols)
+# for col in flcols:
+#     data_train[col] = normcols[col]
+
+X_train = data_train[[1,2,3,4]]
+y_train = data_train[data_train.columns[0]]
+# print(X_train[3].value_counts())
+# print(X_train[4].value_counts())
+#print(X_train)
+
+# train_columns = data_train.columns[5]
+# text_columns = [3,4]
+# data_train_dummy = pd.get_dummies(data_train[columns])
+# print(len(data_train_dummy.columns))
+data_val_X = pd.read_csv("dev-0/in.tsv", header=None, sep='\t')
+data_val_y = pd.read_csv("dev-0/expected.tsv", header=None)
+
+data_val_X.replace({2: brand_codes}, inplace=True)
+data_val_X.replace({3: fuel_codes}, inplace=True)
+#print(data_val_X[2].value_counts())
+#print(data_val_X[3].value_counts())
+X_val = data_val_X[[0,1,2,3]]
+#print(data_val_y)
+reg = LinearRegression().fit(X_train, y_train)
+print(reg.score(X_train, y_train))
+print(reg.score(X_val, data_val_y))
+print(mean_squared_error(data_val_y, reg.predict(X_val), squared=False))
+
+file = open('dev-0/out.tsv',"w") 
+for index, row in X_val.iterrows():
+    #print(np.reshape(row.to_numpy(),(-1,1)))
+    y_pred = reg.predict(np.reshape(row.to_numpy(),(1,-1)))
+    # print(y_pred)
+    file.writelines("{}\n".format(y_pred[0]))
+    # if index==10:
+    #     break  
+file.close()
+
+
+data_test_X = pd.read_csv("test-A/in.tsv", header=None, sep='\t')
+data_test_X.replace({2: brand_codes}, inplace=True)
+data_test_X.replace({3: fuel_codes}, inplace=True)
+data_test_X.replace({2: {"Fabrycznie": 90}}, inplace=True)
+# print(data_test_X[2].value_counts())
+# print(data_test_X[3].value_counts())
+X_test = data_test_X[[0,1,2,3]]
+#print(X_test)
+
+file = open('test-A/out.tsv',"w") 
+for index, row in X_test.iterrows():
+    #print(np.reshape(row.to_numpy(),(-1,1)))
+    y_pred = reg.predict(np.reshape(row.to_numpy(),(1,-1)))
+    # print(y_pred)
+    file.writelines("{}\n".format(y_pred[0]))
+    # if index==10:
+    #     break  
+file.close()
+# #Normalizacja danych
+# flcols = data_val_X[[4]].columns
+# x = data_val_X[[4]].values
+# # min_max_scaler = preprocessing.MinMaxScaler()
+# max_abs_scaler = preprocessing.MaxAbsScaler()
+# # x_scaled = min_max_scaler.fit_transform(x)
+# x_scaled = max_abs_scaler.fit_transform(x)
+# normcols = pd.DataFrame(x_scaled, columns=flcols)
+# for col in flcols:
+#     data_val_X[col] = normcols[col]
+# #Normalizacja danych
+# flcols = data_val_y[[0]].columns
+# x = data_val_y[[0]].values
+# # min_max_scaler = preprocessing.MinMaxScaler()
+# max_abs_scaler = preprocessing.MaxAbsScaler()
+# # x_scaled = min_max_scaler.fit_transform(x)
+# x_scaled = max_abs_scaler.fit_transform(x)
+# normcols = pd.DataFrame(x_scaled, columns=flcols)
+# for col in flcols:
+#     data_val_y[col] = normcols[col]
+
+