Host Dataset externally - plot_boxes

[srinidhi98]

Try to host dataset externally on any web server/cloud and use the URL to extract the data directly into dataFrame/download as a .csv file local repo.

[iingram]

I know that hosting on Kaggle is still amongst the options being considered but I do feel that that meaning there's the extra step for a fresh user of making an account on Kaggle to be able to do the data download is very plausibly a non-negligible barrier that we want to avoid. I am noting that to keep us thinking about it but I am certainly still game for seeing the full side-by-side that includes Kaggle as an option and fully weighing the pros and cons.

[srinidhi98]

For Kaggle, yes the user must have

• account created because Kaggle username is required

• Dataset name

sample code:
from zipfile import ZipFile
import pandas as pd
from kaggle.api.kaggle_api_extended import KaggleApi# Initialize Kaggle API
api = KaggleApi()
api.authenticate()# authenticate the owner's name and dataset
dataset_loc = 'srinidhiyerabati/Test-box-plots'# username/Dataset name on Kaggle
api.dataset_download_files(dataset_loc)# download
zip_file_path = '/home/srinidhiyerbati/Desktop/Srinidhi_Yerabati/animl-py/animl-py/Test-box-plots.zip'

Extract the zip file

with ZipFile(zip_file_path, 'r') as zip_ref:
zip_ref.extractall('/home/srinidhiyerbati/Desktop/Srinidhi_Yerabati/animl-py/animl-py/')
file_path = '/home/srinidhiyerbati/Desktop/Srinidhi_Yerabati/animl-py/animl-py/detections_plotBoxes.csv'
df = pd.read_csv(file_path)
print(df.head())