test

linear_regression.py

@@ -0,0 +1,81 @@
+import numpy as np
+import pandas as pd
+import torch
+import torch.nn as tnn
+
+filedir = '/home/ubuntu/Pulpit/TAU/mieszkania5'
+#train size
+learningRate = 0.0001 
+epochs = 10000
+#treainfile
+trainfile = filedir + '/train/train.tsv'
+#data files
+dev0in = filedir + '/dev-0/in.tsv'
+dev0out = filedir + '/dev-0/out.tsv' 
+testAin = filedir + '/test-A/in.tsv'
+testAout = filedir + '/test-A/out.tsv'
+
+
+class linearRegression(tnn.Module):
+    def __init__(self, dim_i, dim_o):
+        super(linearRegression, self).__init__()
+        self.linear = tnn.Linear(dim_i, dim_o)
+
+    def forward(self, x):
+        out = self.linear(x)
+        return out
+
+model = linearRegression(1, 1)
+device = torch.device('cpu')
+model.to(device)
+print('model regresji gotowy')
+
+#dane do treningu
+trainfile_read = pd.raed_csv(trainfile, sep='\t', header=None, index_col=None)
+train_data_sizes = np.array(trainfile_read[8].tolist(), dtype=np.float32).reshape(-1, 1)
+train_data_prices = np.array(trainfile_read[0].tolist(), dtype=np.float32).reshape(-1, 1)
+
+#dane do przewidywania
+devfile_read = pd.raed_csv(dev0in, sep='\t', header=None, index_col=None)
+testfile_in = pd.raed_csv(testAin, sep='\t', header=None, index_col=None)
+dev_data_sizes = np.array(devfile_read[7].tolist(), dtype=np.float32)
+test_data_sizes = np.array(testfile_in[7].tolist(), dtype=np.float32)
+criterion = tnn.MSELoss() 
+optimizer = torch.optim.SGD(model.parameters(), lr=learningRate)
+
+print('dane wczytane')
+
+#trening
+for epoch in range(epochs):
+    inputs = torch.from_numpy(train_data_sizes)
+    labels = torch.from_numpy(train_data_prices)
+    
+    optimizer.zero_grad()
+
+    outputs = model(inputs)
+
+    loss = criterion(outputs, labels)
+
+    loss.backward()
+
+    optimizer.step()
+print('trening zakonczony')
+#przewidywanie ceny
+
+predicted = model(torch.from_numpy(dev_data_sizes).requires_grad_()).data.numpy()
+
+print('zapisywanie wyników dev-0')
+
+dev_of = open(dev0out, 'w')
+for i in predicted:
+    dev_of.write(str(i[0])+'\n')
+dev_of.close()
+
+predicted = model(torch.from_numpy(test_data_sizes).requires_grad_()).data.numpy()
+
+print('zapisywanie wyników test-A')
+
+test_of = open(testAout, 'w')
+for i in predicted:
+    test_of.write(str(i[0])+'\n')
+test_of.close()