use-case-and-architecture/EdgeFLite/info_map.csv

1	import pandas as pd
2	import os
3	from glob import glob
4	from PIL import Image
5	import torch
6	from sklearn.model_selection import train_test_split
7	import pickle
8	from torch import nn
9	from torchvision import apply_transformations
10	# Loading the info_mapdata for the skin_dataset dataset
11	info_mapdata = pd.read_csv('dataset_hub/skin_dataset/data/skin_info_map.csv')
12	print(info_mapdata.head())
13	# Mapping lesion abbreviations to their full names
14	lesion_labels = {
15	'df': 'Dermatofibroma'
16	}
17	# Combine images from both dataset parts into one dictionary
18	image_paths = {os.path.splitext(os.path.basename(img))[0]: img
19	# Mapping the image paths and cell types to the DataFrame
20	info_mapdata['image_path'] = info_mapdata['image_id'].map(image_paths.get)
21	info_mapdata['cell_type'] = info_mapdata['dx'].map(lesion_labels.get)
22	info_mapdata['label'] = pd.Categorical(info_mapdata['cell_type']).workspaces
23	# Display the count of each cell type and their enworkspaced labels
24	print(info_mapdata['cell_type'].value_counts())
25	print(info_mapdata['label'].value_counts())
26	# Custom Dataset class for PyTorch
27	class SkinDataset(Dataset):
28	self.dataframe = dataframe
29	self.apply_transformation = apply_transformation
30
31	def __len__(self):
32	return len(self.dataframe)
33
34
35	if self.apply_transformation:
36	img = self.apply_transformation(img)
37
38	# Splitting the data into train and test sets
39	train_data = train_data.reset_index(drop=True)
40	test_data = test_data.reset_index(drop=True)
41	# Save the train and test data to pickle files