How to read a mounted datafile within a python jupyter notebook

[Varnikaa]

Hi,

I have uploaded the data and running the job using the below command:

floyd run --cpu --data varnikaa/datasets/cp_train/1:/train --mode jupyter

I dont know how to load this data into jupyter notebook. Something like this in pandas:

import pandas as pd
train_df = pd.read_csv('train.csv', nrows = 10)

However, I could see that the data has been loaded successfully.

!ls /train
train.csv

I dont know how to read this data into a pandas dataframe.

If possible, please help.

[redeipirati]

Hi @Varnikaa,

With --data varnikaa/datasets/cp_train/1:/train you have mounted your dataset inside /train folder inside your job, so if your notebook runs in a different folder(default workdir is /output), you need to specify the absolute path to retrieve your dataset file: train_df = pd.read_csv('/train/train.csv', nrows = 10).

Hope it helps.

[sai]

Here's the complete documentation on mounting datasets: https://docs.floydhub.com/guides/data/mounting_data/

For specific example on how to use the mounted data in Jupyter Notebook, see this.

[Varnikaa]

Hi @redeipirati,

Thank you very much, This has worked for me