ium_434732/mongoObserver.py

import torch
from torch import nn
import numpy as np
import pandas as pd
from sklearn.metrics import accuracy_score
from sklearn.metrics import f1_score
from sacred import Experiment
from sacred.observers import MongoObserver

np.set_printoptions(suppress=False)

ex = Experiment("434732-mongo", interactive=False, save_git_info=False)
ex.observers.append(MongoObserver(url='mongodb://mongo_user:mongo_password_IUM_2021@172.17.0.1:27017', db_name='sacred'))

@ex.config
def my_config():
    epochs = 5
    batch_size = 10

class LogisticRegressionModel(nn.Module):
    def __init__(self, input_dim, output_dim):
        super(LogisticRegressionModel, self).__init__()
        self.linear = nn.Linear(input_dim, output_dim)
        self.sigmoid = nn.Sigmoid()
    def forward(self, x):
        out = self.linear(x)
        return self.sigmoid(out)


@ex.capture
def readAndtrain(epochs, batch_size, _run):
    train = pd.read_csv("train.csv")
    test = pd.read_csv("test.csv")

    xtrain = train[['age','anaemia','creatinine_phosphokinase','diabetes', 'ejection_fraction', 'high_blood_pressure', 'platelets', 'serum_creatinine', 'serum_sodium', 'sex', 'smoking']].astype(np.float32)
    ytrain = train['DEATH_EVENT'].astype(np.float32)
    xtest = test[['age','anaemia','creatinine_phosphokinase','diabetes', 'ejection_fraction', 'high_blood_pressure', 'platelets', 'serum_creatinine', 'serum_sodium', 'sex', 'smoking']].astype(np.float32)
    ytest = test['DEATH_EVENT'].astype(np.float32)
    xTrain = torch.from_numpy(xtrain.values)
    yTrain = torch.from_numpy(ytrain.values.reshape(179,1))
    xTest = torch.from_numpy(xtest.values)
    yTest = torch.from_numpy(ytest.values)

    learning_rate = 0.002
    input_dim = 11
    output_dim = 1

    _run.log_scalar("Batch", str(batch_size))
    _run.log_scalar("epoch",  str(epochs))

    model = LogisticRegressionModel(input_dim, output_dim)
    model.load_state_dict(torch.load('DEATH_EVENT.pth'))

    criterion = torch.nn.BCELoss(reduction='mean')
    optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)

    for epoch in range(epochs):
        # print ("Epoch #",epoch)
        model.train()
        optimizer.zero_grad()
        # Forward pass
        y_pred = model(xTrain)
        # Compute Loss
        loss = criterion(y_pred, yTrain)
        # print(loss.item())
        # Backward pass
        loss.backward()
        optimizer.step()

    _run.log_scalar("Lost", str(loss.item()))


    torch.save(model.state_dict(), 'DEATH_EVENT.pth')
    prediction= model(xTest)
    _run.log_scalar("accuracy_score", accuracy_score(yTest, np.argmax(prediction.detach().numpy(), axis=1)))
    # _run.log_scalar("F1", str(f1_score(yTest, np.argmax(prediction.detach().numpy(), axis=1), average=None)))

    print("accuracy_score", accuracy_score(yTest, np.argmax(prediction.detach().numpy(), axis=1)))
    # print("F1", f1_score(yTest, np.argmax(prediction.detach().numpy(), axis=1), average=None))

@ex.automain
def my_main(epochs, batch_size, _run):
    readAndtrain()

ex.run()
ex.add_artifact('DEATH_EVENT.pth')