/pandas-hacks

Repo showing some neat packages for Pandas

Primary LanguageJupyter NotebookMIT LicenseMIT

pandas-hacks

Repo showing some neat packages for Pandas

Quickstart

  1. Install Anaconda
  2. Install direnv (Optional)
  3. Clone this repo: git clone https://github.com/Tommo565/pandas-hacks.git and cd into the repo cd pandas-hacks
  4. Create a service account with BigQuery > Editor permissions. This will allow Python and Jupyter to interact with Bigquery.
  5. Create a service account key and save this somewhere outside of the repo.
  6. If using direnv, create an .envrc file and a GOOGLE_APPLICATION_CREDENTIALS variable with the path to your credentials file. You should be prompted to execute direnv 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"
  1. 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"
  1. Create the conda environment conda env create -f env.yml
  2. Install the environment into Jupyter python -m ipykernel install --name pandas-hacks
  3. Start the Jupyter notebook jupyter notebook

Contents

The notebooks in the notebooks folder contain overviews for the following pandas packages:

  • pandas-gbq
  • pandas-profiling
  • dtale