Materials for my PyData NYC 2017 tutorail

1. Install Miniconda for your platform (instructions)

• If you already have conda, you can skip this step; It's also fine to use virtualenv + pip. You should be able to pip install -r requirements.txt to get the same package versions as step 4.
• Make sure to open a new terminal shell after installing, so that conda is on your path

2. Clone the repository at https://github.com/tomaugspurger/pydata-nyc-ph2t

• git clone https://github.com/tomaugspurger/pydata-nyc-ph2t
• If you don't have git installed, you can download the zip using the green "Clone or download" button, and then "Download ZIP". Note that the filename will be "pydata-nyc-ph2t-master"

3. Change into the repository

• cd pydata-nyc-ph2t

4. Create the conda environment

• conda env create

5. Activate the environment

• conda activate ph2t if you have a new enough conda
• source activate ph2t on Linux or MacOS
• activate ph2t on Windows

6. Start the Jupyter notebook server

• jupyter notebook

Once your notebook server is running, (jupyter notebook) your browser should open up to the webpage (http://localhost:8888 by default). Open the notebook notebooks/00-README.ipynb and familiarize yourself with using notebook.

conda create -n ph2t python=3.6
# activate / source activate ph2t
conda install numpy=1.13 pandas=0.21.0 matplotlib=2.1 seaborn=0.8 ipython=6.2 jupyter=1.0.0 notebook=5.2 dask=0.15.4 distributed=1.19 toolz=0.8 pandas-datareader=0.5 scikit-learn=0.19 scipy=0.19 statsmodels=0.8 pyarrow=0.7.1 -c conda-forge
pip install lifetime altair