How to calculate the f1-score? - pytorch

Question

I have a pyTorch-code to train a model that should be able to detect placeholder-images among product-images. I didn't write the code by myself as I am very unexperienced with CNNs and Machine Learning.
My boss told me to calculate the f1-score for that model and i found out that the formula for that is ((precision * recall)/(precision + recall)) but I don't know how I get precision and recall. Is someone able to tell me how I can get those two parameters from that following code?
(Sorry for the long piece of code, but I didn't really know what is necessary and what isn't)
from __future__ import print_function
from __future__ import division
import torch
import torch.nn as nn
import torch.optim as optim
import numpy as np
import torchvision
from torchvision import datasets, models, transforms
import matplotlib.pyplot as plt
import time
import os
import copy
print("PyTorch Version: ",torch.__version__)
print("Torchvision Version: ",torchvision.__version__)
data_dir = "data"
# Models to choose from [resnet, alexnet, vgg, squeezenet, densenet, inception]
model_name = "resnet"
# Number of classes in the dataset [we have four classes A-Balik-Duz-Princess]
num_classes = 2
# Batch size for training (change depending on how much memory you have)
batch_size = 25
# Number of epochs to train for (This will need to be calculated in order to address under and over fitting issue)
num_epochs = 20
# Flag for feature extracting. When False, we fine tune the whole model,
# when True we only update the reshaped layer params
feature_extract = True
def train_model(model, dataloaders, criterion, optimizer, num_epochs=25, is_inception=False):
since = time.time()
print("model is : ",model)
val_acc_history = []
val_loss_history = []
train_acc_history = []
train_loss_history = []
best_model_wts = copy.deepcopy(model.state_dict())
best_acc = 0.0
for epoch in range(num_epochs):
print('Epoch {}/{}'.format(epoch, num_epochs - 1))
print('-' * 10)
# Each epoch has a training and validation phase
for phase in ['train', 'val']:
if phase == 'train':
model.train() # Set model to training mode
else:
model.eval() # Set model to evaluate mode
running_loss = 0.0
running_corrects = 0
# Iterate over data.
for inputs, labels in dataloaders[phase]:
inputs = inputs.to(device)
labels = labels.to(device)
# zero the parameter gradients (This can be changed to the Adam and other optimizers)
optimizer.zero_grad()
# forward
# track history if only in train
with torch.set_grad_enabled(phase == 'train'):
# Get model outputs and calculate loss
# Special case for inception because in training it has an auxiliary output. In train
# mode we calculate the loss by summing the final output and the auxiliary output
# but in testing we only consider the final output.
if is_inception and phase == 'train':
# From https://discuss.pytorch.org/t/how-to-optimize-inception-model-with-auxiliary-classifiers/7958
outputs, aux_outputs = model(inputs)
loss1 = criterion(outputs, labels)
loss2 = criterion(aux_outputs, labels)
loss = loss1 + 0.4*loss2
else:
outputs = model(inputs)
loss = criterion(outputs, labels)
_, preds = torch.max(outputs, 1)
# backward + optimize only if in training phase
if phase == 'train':
loss.backward()
optimizer.step()
# statistics
running_loss += loss.item() * inputs.size(0)
running_corrects += torch.sum(preds == labels.data)
epoch_loss = running_loss / len(dataloaders[phase].dataset)
epoch_acc = running_corrects.double() / len(dataloaders[phase].dataset)
print('{} Loss: {:.4f} Acc: {:.4f}'.format(phase, epoch_loss, epoch_acc))
# deep copy the model
if phase == 'val' and epoch_acc > best_acc:
best_acc = epoch_acc
best_model_wts = copy.deepcopy(model.state_dict())
if phase == 'val':
val_acc_history.append(epoch_acc)
val_loss_history.append(epoch_loss)
if phase == 'train':
train_acc_history.append(epoch_acc)
train_loss_history.append(epoch_loss)
print()
time_elapsed = time.time() - since
print('Training complete in {:.0f}m {:.0f}s'.format(time_elapsed // 60, time_elapsed % 60))
print('Best val Acc: {:4f}'.format(best_acc))
# load best model weights
model.load_state_dict(best_model_wts)
return model, val_acc_history, train_acc_history,val_loss_history,train_loss_history
def set_parameter_requires_grad(model, feature_extracting):
if feature_extracting:
for param in model.parameters():
param.requires_grad = False
###############################################
### Initialize and Reshape the Networks
###############################################
def initialize_model(model_name, num_classes, feature_extract, use_pretrained=True):
# Initialize these variables which will be set in this if statement. Each of these
# variables is model specific.
model_ft = None
input_size = 0
if model_name == "resnet":
""" Resnet18
"""
model_ft = models.resnet152(pretrained=use_pretrained)
#we can select any possible variation of ResNet such as Resnet18, Resnet34, Resnet50, Resnet101, and Resnet152
set_parameter_requires_grad(model_ft, feature_extract)
num_ftrs = model_ft.fc.in_features
model_ft.fc = nn.Linear(num_ftrs, num_classes)
input_size = 224
elif model_name == "alexnet":
""" Alexnet
"""
model_ft = models.alexnet(pretrained=use_pretrained)
set_parameter_requires_grad(model_ft, feature_extract)
num_ftrs = model_ft.classifier[6].in_features
model_ft.classifier[6] = nn.Linear(num_ftrs,num_classes)
input_size = 224
elif model_name == "vgg":
""" VGG11_bn
"""
model_ft = models.vgg11_bn(pretrained=use_pretrained)
set_parameter_requires_grad(model_ft, feature_extract)
num_ftrs = model_ft.classifier[6].in_features
model_ft.classifier[6] = nn.Linear(num_ftrs,num_classes)
input_size = 224
elif model_name == "squeezenet":
""" Squeezenet
"""
model_ft = models.squeezenet1_0(pretrained=use_pretrained)
set_parameter_requires_grad(model_ft, feature_extract)
model_ft.classifier[1] = nn.Conv2d(512, num_classes, kernel_size=(1,1), stride=(1,1))
model_ft.num_classes = num_classes
input_size = 224
elif model_name == "densenet":
""" Densenet
"""
model_ft = models.densenet121(pretrained=use_pretrained)
set_parameter_requires_grad(model_ft, feature_extract)
num_ftrs = model_ft.classifier.in_features
model_ft.classifier = nn.Linear(num_ftrs, num_classes)
input_size = 224
elif model_name == "inception":
""" Inception v3
Be careful, expects (299,299) sized images and has auxiliary output
"""
model_ft = models.inception_v3(pretrained=use_pretrained)
set_parameter_requires_grad(model_ft, feature_extract)
# Handle the auxilary net
num_ftrs = model_ft.AuxLogits.fc.in_features
model_ft.AuxLogits.fc = nn.Linear(num_ftrs, num_classes)
# Handle the primary net
num_ftrs = model_ft.fc.in_features
model_ft.fc = nn.Linear(num_ftrs,num_classes)
input_size = 299
else:
print("Invalid model name, exiting...")
exit()
return model_ft, input_size
# Initialize the model for this run
model_ft, input_size = initialize_model(model_name, num_classes, feature_extract, use_pretrained=True)
# Print the model we just instantiated
#print(model_ft)
########################
### LOAD DATA
########################
# Data augmentation and normalization for training
# there are multiple approaches for data augmentation which can be added in the future
# Just normalization for validation
data_transforms = {
'train': transforms.Compose([
transforms.RandomResizedCrop(input_size),
#transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
]),
'val': transforms.Compose([
transforms.Resize(input_size),
transforms.CenterCrop(input_size),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
]),
}
print("Initializing Datasets and Dataloaders...")
# Create training and validation datasets
image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x), data_transforms[x]) for x in ['train', 'val']}
# Create training and validation dataloaders
dataloaders_dict = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=batch_size, shuffle=True, num_workers=4) for x in ['train', 'val']}
# Detect if we have a GPU available
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
#############################
### Create the Optimizer
#############################
# Send the model to GPU
model_ft = model_ft.to(device)
# Gather the parameters to be optimized/updated in this run. If we are
# fine tuning we will be updating all parameters. However, if we are
# doing feature extract method, we will only update the parameters
# that we have just initialized, i.e. the parameters with requires_grad
# is True.
params_to_update = model_ft.parameters()
print("Params to learn:")
if feature_extract:
params_to_update = []
for name,param in model_ft.named_parameters():
if param.requires_grad == True:
params_to_update.append(param)
print("\t",name)
else:
for name,param in model_ft.named_parameters():
if param.requires_grad == True:
print("\t",name)
# Observe that all parameters are being optimized we can add leaky ReLU and much more
optimizer_ft = optim.SGD(params_to_update, lr=0.001, momentum=0.9)
###########################
### Run Training and Validation Step
###########################
%time
# Setup the loss fxn
criterion = nn.CrossEntropyLoss()
# Train and evaluate
model_ft, hist, loss_t,vloss_acc, tloss_acc = train_model(model_ft, dataloaders_dict, criterion, optimizer_ft, num_epochs=num_epochs, is_inception=(model_name=="inception"))

Answer 1

...
# statistics
running_loss += loss.item() * inputs.size(0)
running_corrects += torch.sum(preds == labels.data)
# Add these lines to obtain f1_score
from sklearn.metrics import f1_score
f1_score = f1_score(labels.data, preds)
#or: f1_score = f1_score(labels.cpu().data, preds.cpu())
...

Answer 2

You can use sklearn to calculate f1_score
from sklearn.metrics import f1_score
X, y = get_data(...)
y_pred = model.predict(X)
f1_score(y, y_pred)

Answer 3

I think it's better to call f1-score with macro/micro.
from sklearn.metrics import f1_score
print('F1-Score macro: ',f1_score(outputs, labels, average='macro'))
print('F1-Score micro: ',f1_score(outputs, labels, average='micro'))
The key difference between micro and macro F1 score is their behavior on imbalanced datasets. Micro F1 score often doesn't return an objective measure of model performance when the classes are imbalanced, whilst Macro F1 score is able to do so.Read More