Citibike Machine Learning Hands-on-Lab with Snowpark Python

Requirements:

-An account in Amazon SageMaker Studio Lab. Do this ahead of time as sign-up may take up to 24 hours due to backlog. -A Snowflake account activated for Snowpark Private Preview.

Setup Steps:

-Login to SageMaker Studio Lab
-Create a Runtime if there isn't one already
-Click on Start Runtime
-Click on Open Project
-Select Git -> Clone Git Repository and enter the following:
-- Repository URL: https://github.com/sfc-gh-mgregory/snowpark-python-hol.
-Select Yes when prompted to create a conda environment. -When opening notebooks be sure to select the "snowpark_041" kernel.

Alternative Client

As an alternative to SageMaker Studio Lab this hands-on-lab can be run in Jupyter from local systems. The following example is for MacOS.

curl https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.sh -o ~/Downloads/miniconda.sh  
sh ~/Downloads/miniconda.sh -b -p $HOME/miniconda  
~/miniconda/bin/conda init  
conda update conda
cat .bash_profile >> .zshrc  

-If git is not installed on your local system you can install via conda.

conda install git

-Clone this repository and create an environment

mkdir ~/Desktop/snowpark-python
cd ~/Desktop/snowpark-python
git clone https://github.com/sfc-gh-mgregory/snowpark-python-hol
cd snowpark-python-hol
conda env create -f environment.yml
conda activate snowpark_041
jupyter notebook

TODO

-Update with public conda install of snowpark client