First draft of project start script (WIP)

createproject.sh

@@ -0,0 +1,94 @@
+# This is meant to create a Data Science project following the
+# format here: 
+# https://github.com/drivendata/cookiecutter-data-science
+# with some minor tweaks and with some automation.
+
+# Some updates in the drivendata repo are expected soon, which will
+# require an update from this (Apr 12, 2019).
+
+# -------------------------------------------------------------------
+# REQUIREMENTS
+# -------------------------------------------------------------------
+# conda/miniconda
+# conda-forge channel added. (conda config --add channels conda-forge)
+# 
+# -------------------------------------------------------------------
+# HOW TO USE
+# -------------------------------------------------------------------
+# Go to a parent folder where your repo folder will go in. Then run
+# ./path/to/createproject.sh
+# -------------------------------------------------------------------
+
+# First we create a conda environment.
+echo "First we create a conda environment for our new project."
+read -p "Enter the conda env name (recommended same name as repo) : " environment_name
+conda create -n $environment_name
+# remove is conda remove --name myenv --all
+conda activate $environment_name
+
+# Check if this works without adding the conda-forge channel to the base env
+echo "Then we install cookiectter."
+conda install cookiecutter
+
+# Packages to install from conda
+echo "Let's install some conda packages."
+read -p "List the packages you want to install from conda : " condapkg_install
+if [[ ! -z "$condapkg_install" ]]; then
+	conda install $condapkg_install
+fi
+
+# echo "Now for pip packages. Only pick packages you can't get from conda."
+# echo "I recommend you install dotenv. It keps you manage local environment variables and API keys."
+# echo "I also recommend you install kaggle if you want to download a dataset from there."
+# read -p "List the packages you want to install from pip : " pippkg_install
+# pip install $pippkg_install
+
+# echo "Please name your repo/project the same as the conda env name or else this script will break."
+# cookiecutter https://github.com/drivendata/cookiecutter-data-science
+
+# cd $environment_name
+
+# Download data from kaggle
+# cd data/raw
+# if [[ $pippkg_install == *"kaggle"* ]]; then
+# 	echo "You have opted to install kaggle. If you want to download data, make sure"
+# 	echo "you have set up your API key as described in https://github.com/kaggle/kaggle-api"
+# 	read -p "Do you want to download a dataset? [y/n] : " kag_dwn
+# 	if [[ $kag_dwn == "y" ]]; then
+# 		echo "Enter the link to the dataset. For example, if "
+# 		echo "the link to the dataset is kaggle.com/mlg-ulb/creditcardfraud "
+# 	        echo "then enter mlg-ulb/creditcardfraud"
+# 		read -p "Enter link : " kag_das
+# 		kaggle datasets download -d $kag_das
+# 	fi
+# fi
+# cd ../..
+
+# Replace the data folder with nothing.
+# sed -i -e "s_/data/_#Nothing" .gitignore
+
+# Tweak .gitignore file.
+# echo "# Include data sub-folders" >> .gitignore
+# echo "data/raw/*" >> .gitignore
+# echo "data/interim/*" >> .gitignore
+# echo "data/processed/*" >> .gitignore
+# echo "data/external/*" >> .gitignore
+# echo "!.gitkeep" >> .gitignore
+
+# Replace .env with *.env
+# sed -i -e 's/.env/*.env' .gitignore
+
+# Rename package from src to something else.
+# echo "Your package will be given the same name as your conda env/repo"
+# echo "Except any non-letters will be taken out."
+# ppkg_name=$(echo $environment_name | sed 's/[^a-zA-Z]*//g')
+# mv src $ppkg_name
+# sed -i -e "s/src/$ppkg_name/g" setup.py
+
+# Pip install package.
+# pip install -e .
+
+# Initialize and push git repo.
+
+# That's it for now!
+