Access BQ from Python

You need to install pandas-gbq first.

pip install pandas-gbq

If you’re using Google Colab

It’s easy. Authenticate to all Google cloud services your user-account has access to in this way:

from google.colab import auth
auth.authenticate_user()

Once that’s done, run your query from pandas-gbq.

import pandas as pd
PROJECT_ID = ""
df = pd.read_gbq(query,project_id=PROJECT_ID)

If you’re using a VM

Using Service Account

Follow the steps here to create a service account.

from google.oauth2 import service_account
import pandas_gbq

credentials = service_account.Credentials.from_service_account_file(
    'path/to/key.json',
)
df = pandas_gbq.read_gbq(sql, project_id="YOUR-PROJECT-ID", credentials=credentials)

Using a user-account

Install the gcloud CLI from here.

Run the following to authenticate your user account.

gcloud auth login

You need to install pydata_google_auth as well.

pip install pydata-google-auth

Then you can access BQ from your notebook or script in this way:

import pandas_gbq
import pydata_google_auth
credentials, project = pydata_google_auth.default()
PROJECT_ID = ""

df = pd.read_gbq(query,project_id = PROJECT_ID, credentials = credentials)