sjsrey / pysalnarsc17

PySAL Workshop at NARSC17

Spatial Data Analysis with PySAL @NARSC2017

Instructors

• Sergio Rey - University of California, Riverside
• Wei Kang - Arizona State University

This repository contains the materials and instructions for the PySAL workshop at NARSC 2017.

Schedule

• 8:00-10:00
  • Overview of PySAL and workshop
  • Introductions
  • Installation
  • Jupyter notebooks
  • Python primer
• 10:00-10:30
  • Coffee Break
• 10:30-12:30
  • Spatial data processing
  • Choropleth mapping and geovisualization
  • Spatial weights
• 12:30-1:30
  • Lunch
• 1:30-3:00
  • Global spatial autocorrelation
  • Local spatial autocorrelation
  • Spatial inequality analysis
• 3:00-3:30
  • Coffee Break
• 3:30-5:00
  • Geodemographics and regionalization
  • Spatial dynamics
  • Spatial regression

Obtaining Workshop Materials

If you are familiar with GitHub, you should clone or fork this GitHub repository to a specific directory. Cloning can be done by:

git clone https://github.com/sjsrey/pysalnarsc17.git

If you are not using git, you can grab the workshop materials as a zip file by pointing your browser to (https://github.com/sjsrey/pysalnarsc17.git) and clicking on the green Clone or download button in the upper right.

Extract the downloaded zip file to a working directory.

Installation

We will be using a number of Python packages for geospatial analysis.

An easy way to install all of these packages is to use a Python distribution such as Anaconda. In this workshop we will be using Python 2.7 so please download that version of Anaconda.

Once you have downloaded Anaconda, start a terminal and navigate to the directory of the downloaded/ cloned materials. For example, if the materials now live in the directory /Users/weikang/Downloads/pysalnarsc17-master, you need to navigate to that directory from the terminal (using command cd):

Once we have done that, run:

conda-env create -f workshop.yml

This will build a conda environment that sandboxes the installation of the required packages for this workshop so we don't break anything in your computer's system Python (if it has one).

This may take 10-15 minutes to complete depending on the speed of your network connection.

Once this completes, you can activate the workshop environment with:

• on Mac, Linux

source activate workshop

• on Windows:

activate workshop

Next, you will want to test your installation with:

 jupyter-nbconvert --execute --ExecutePreprocessor.timeout=120 check_workshop.ipynb

You should see something like:

[NbConvertApp] Converting notebook check_workshop.ipynb to html
[NbConvertApp] Executing notebook with kernel: python2
[NbConvertApp] Writing 435635 bytes to check_workshop.html

Open check_workshop.html in a browser, and scroll all the way down, you should see something like:

If you do see the above, you are ready for the tutorial. If not, please contact either of us for help.

pysalnarsc17's Issues

datashader import issue in 18_points

The reason is that conda-forge chanel does not give the most updated version of datashader. It is version 0.3.0 on conda-forge channel.

Need to add the channel "bokeh" to get the most updated version of datashader (0.6.2) to get it to work.

lon and lat are wrong in participants.csv

-180<= lon <= 180

-90 <= lat <= 90

folium example is broken in 09_geovisualization

boken missing charts in 09_geovisualization

SPVCM example appended to the spatial regression notebook

This will not be ready by the time the workshop is due (eod today in north america), but I figured I'd file here so that the idea & the intent was tracked.

Jupyter cell types

In 00_notebook_intro.ipynb, the second cell type should be raw cell instead of text cell. The keyboard shortcut should be R instead of t.

update install instructions

simplify these to point to a terminal (agnostic to os) and use scheme similar to that used at scipy for sklearn tutorials