Repo showing some neat packages for Pandas
- Install Anaconda
- Install direnv (Optional)
- Clone this repo:
git clone https://github.com/Tommo565/pandas-hacks.git
andcd
into the repocd pandas-hacks
- Create a service account with BigQuery > Editor permissions. This will allow Python and Jupyter to interact with Bigquery.
- Create a service account key and save this somewhere outside of the repo.
- If using direnv, create an
.envrc
file and aGOOGLE_APPLICATION_CREDENTIALS
variable with the path to your credentials file. You should be prompted to executedirenv allow
in the command line. This will load the variable into your environment automatically. The file should look like this:
export GOOGLE_APPLICATION_CREDENTIALS="path/to/your/service_account_key.json"
- If not using direnv, you will need a create a GOOGLE_APPLICATION_CREDENTIALS variable in the notebook as follows:
import os
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "path/to/your/service_account_key.json"
- Create the conda environment
conda env create -f env.yml
- Install the environment into Jupyter
python -m ipykernel install --name pandas-hacks
- Start the Jupyter notebook
jupyter notebook
The notebooks in the notebooks
folder contain overviews for the following pandas packages:
- pandas-gbq
- pandas-profiling
- dtale