
!conda install -qq -y -c conda-forge gdown

Collecting package metadata (current_repodata.json): ...working... done
Solving environment: ...working... done

# All requested packages already installed.


#@title Download Dataset
# https://drive.google.com/file/d/1JWfTkCuYCaOx0PKrjUZAks87DKPQSezM/view?usp=sharing
!gdown 1JWfTkCuYCaOx0PKrjUZAks87DKPQSezM
!unzip -q *.zip
!rm *.zip
!rm -rf __MACOSX/
!mv 1DCA-ME88/* /home/ec2-user/SageMaker/
!rm -rf 1DCA-ME88
!rm -rf sample_data/


import csv

# Open the input file
with open('master_list_of_files.csv', 'r') as input_file:
    reader = csv.reader(input_file)

    # Read the data rows
    data = [row for row in reader]

# Add the header
header = ['path1', 'path2', 'path3']
data_with_header = [header] + data

# Write the data with header to a new CSV file
with open('paths.csv', 'w', newline='') as output_file:
    writer = csv.writer(output_file)
    writer.writerows(data_with_header)

# print("Header added successfully!")

!rm master_list_of_files.csv

Downloading...
From (uriginal): https://drive.google.com/uc?id=1JWfTkCuYCaOx0PKrjUZAks87DKPQSezM
From (redirected): https://drive.google.com/uc?id=1JWfTkCuYCaOx0PKrjUZAks87DKPQSezM&confirm=t&uuid=ae74a28a-4ab8-46e3-adcd-97549444d8c4
To: /home/ec2-user/SageMaker/1DCA-ME88.zip
100%|███████████████████████████████████████| 27.3M/27.3M [00:00<00:00, 173MB/s]


#@title Sort csv file


import csv

def sort_csv(filename):
    # Read the CSV file into a list of rows
    with open(filename, 'r') as file:
        reader = csv.reader(file)
        rows = list(reader)

    # Sort the rows based on the first column, number1, and descending float number3
    sorted_rows = sorted(rows[1:], key=lambda row: (int(row[0].split('_')[1]), int(row[0].split('_')[3]), -float(row[0].split('_')[5][:-4])))

    # Write the sorted rows back to the CSV file
    with open("sorted_paths.csv", 'w', newline='') as file:
        writer = csv.writer(file)
        writer.writerows([rows[0]] + sorted_rows)

# Usage example
filename = 'paths.csv'
sort_csv(filename)
print(f"The CSV file '{filename}' has been sorted successfully.")

The CSV file 'paths.csv' has been sorted successfully.


# @title Slice the csv for a range because you would not be able to render all images in one time (Commented code, use if required)

import csv

def select_rows(input_file, output_file, start_row, end_row):
    with open(input_file, 'r', newline='') as f_in:
        reader = csv.reader(f_in)
        rows = list(reader)

        headers = rows[0]
        selected_rows = rows[start_row-1:end_row]

    with open(output_file, 'w', newline='') as f_out:
        writer = csv.writer(f_out)
        writer.writerow(headers)
        writer.writerows(selected_rows)

# Specify the range of rows you want to select
m = 500  # Start row
n = 1000  # End row

input_file = 'sorted_paths.csv'
output_file = f'sorted_paths_{m}_{n}.csv'

select_rows(input_file, output_file, m, n)


#@title Display all results


import pandas as pd
from IPython.display import display, HTML
import base64
# Read the CSV file
df = pd.read_csv(output_file)

# Apply CSS style for zooming
display(HTML('''
<style>
.zoomable-image {
  transition: transform 0.2s;
}

.zoomable-image:hover {
  transform: scale(1.5);
}
</style>
'''))

# Iterate over each row in the DataFrame
for _, row in df.iterrows():
    html_string = ''
    html_string += '<div style="display: flex; flex-direction: row; margin: 10px;">'

    # Iterate over each image path in the row
    for col in ['path3', 'path2', 'path1']:
        image_path = row[col]
        image_data = base64.b64encode(open(image_path, 'rb').read()).decode('utf-8')

        html_string += '<div style="display: flex; flex-direction: column; text-align: center; margin-right: 10px;">'
        html_string += f'<p>Image: {image_path}</p>'

        # Add CSS class 'zoomable-image' to path2 image
        if col == 'path2':
            html_string += f'<img src="data:image/svg+xml;base64,{image_data}" class="zoomable-image" width="400" height="400" />'
        else:
            html_string += f'<img src="data:image/svg+xml;base64,{image_data}" width="400" height="400" />'

        html_string += '</div>'

    html_string += '</div>'

    display(HTML(html_string))