Importing CSV file into Google Colab using numpy loadtxt

Question

I'm trying to migrate a JupyterLab notebook to Google Colab. In JupyterLab, when I have the notebook file and the associated csv files in the same directory, it is easy to import the data using numpy's loadtxt function as follows:

import numpy as np
filein = "testfile.csv"
data = np.loadtxt(open(filein, "rb"), delimiter=",", skiprows=1)

For various reasons I'd like to continue to use np.loadtxt in Colab. However when I try the same code there, it can't find the csv file despite it residing in the same Google Drive location as the notebook file. I get this error: "FileNotFoundError: [Errno 2] No such file or directory: 'testfile.csv'".

I gather I somehow need to provide a path to the file, but haven't been able to figure out how to do this. Is there any straightforward way to use np.loadtxt?

Answer 1

Colab doesn't automatically mount Google Drive. By default, the working directory is /content on an ephemeral backend virtual machine.

To access your file in Drive, you'll need to mount it first using the following snippet:

from google.colab import drive
drive.mount('/content/gdrive')

Then, %cd /content/gdrive/My\ Drive to change the working directory to your Drive root. (Or, customize the path as needed to wherever testfile.csv is located.)

Answer 2

Shorter and without command

# mount gdrive with this code
from google.colab import drive
drive.mount('/content/drive')
#below where the file is in gdrive, change with your
data_path = "/content/drive/My Drive/Colab Notebooks/test/"
yearsBase, meanBase = np.loadtxt(data_path + 'file.csv', delimiter=',', unpack=True)

done, no other code needed ciao