Keras-tutorial

This repository contains material for a tensorflor-keras tutorial, part of the "Deep learning for climate modeling seminar".

Setup TensorFlow Keras environment

We first need a tensorflow/keras development environment. These steps walk you through setting up an environment using Python 3.6.

1. First we need to create a new environment conda create --name keras
2. Now we need to activate our environment source activate keras
3. Now let’s import all the modules we will use:
   1. This is so we can work with NetCDF files conda install -c conda-forge xarray
   2. This is so we can make plots conda install -c conda-forge matplotlib
   3. This is so we can use TensorFlow and Keras conda install -c conda-forge tensorflow
   4. This is a nice machine learning library. conda install -c conda-forge scikit-learn
   5. This is so our environment can be a Jupyter kernel conda install ipykernel
4. Now make your new environment a Jupyter kernel: python -m ipykernel install --user --name keras --display-name "keras"
5. Now open a Jupyter notebook and make sure you can import everything. You may need to change to your new keras kernel:

import xarray as xr
import numpy as np
import sklearn as sk
import tensorflow as tf
from tensorflow import keras

notebooks

This is a collection of examples using the keras API. Notebooks can be accessed and modified via binder here

doc

This contains documentation files:

• setting_up_a_keras_environment.md
• neural-network-theory.md