Setting Environment Variables

For convenience, I have included a check in my example scripts for an environment variable called “DS_WORKSPACE” that is used for reading from and saving to a designated workspace folder on your computer. If this variable does not exist, the scripts assume you want to read and write to the working folder (i.e., the same folder that your script lives in). To create and define your workspace folder, please follow these instructions.

On Windows

  • On your keyboard, type “Win + s” to open the search bar.
  • Type “environ” in the search bar and select “Edit environment variables for your account”
  • Under “User variables for ‘user’”, click “New…”
  • For Variable name, type DS_WORKSPACE
  • For Variable value, type an absolute path to your directory (e.g., “C:\Workspace”); you may use the “Browse Directory…” button to navigate to your folder of choice.
  • Click OK and click OK again.

On macOS / Linux

Run the following command in the terminal or save it to your ~/.bash_profile to run it every time you open a terminal.

  • export DS_WORKSPACE=~/Workspace

Installing Wheel Files

Python wheel files (.whl) are the new distribution standard for Python packages. You may come across wheel files (especially those of you trying to install numpy on Windows) because they can come with precompiled C code, which you would otherwise have to compile on your machine. You can install a .whl package using pip.

For example, download a wheel package (“package-0.10.2-cp37-cp37m-win_amd64.whl”), open a terminal and cd to folder where it was downloaded, and run pip install package-0.10.2-cp37-cp37m-win_amd64.whl.

Linking Images to Markdown Files

In order for images to appear on your web page that is rendered from markdown, you need a publicly accessible URL for the image.

Two popular cloud-based storage options are Dropbox and Google Drive. You may also find public domain or CC-BY-SA images on Wikimedia Commons. If you want to access images that are saved in one of these locations, try the steps outlined below.

Dropbox

  1. Create a shareable link to a media file (e.g., https://www.dropbox.com/s/pqtsip7582dc3bk/logo.png?dl=0).
  2. Change the ?dl=0 to ?raw=1
![Example link to image stored on Dropbox.](https://www.dropbox.com/s/pqtsip7582dc3bk/logo.png?raw=1){width=150 height=150}

Example link to image stored on Dropbox.

Drive

  1. Create a shareable link to a media file (e.g., https://drive.google.com/file/d/11YjM3ua8DXZvhhfEFaG1LPUupnoj2qB4/view?usp=sharing)
  2. Identify the file ID. It appears after the /file/d and before the /view? in the URL (i.e., 11YjM3ua8DXZvhhfEFaG1LPUupnoj2qB4 from the example URL above).
  3. Take the file ID and place it in a URL formatted as: https://drive.google.com/uc?export=view&id= and append the file ID at the end.
![Example link to image stored in Drive](https://drive.google.com/uc?export=view&id=11YjM3ua8DXZvhhfEFaG1LPUupnoj2qB4){width=150 height=150}

Example link to image stored in Drive

Wikimedia

  1. Find an image you want to link to on https://commons.wikimedia.com
  2. Notice that the preview image page is not the actual image and will not work in markdown. Check out the difference between these two links (yes, both URLs appear to end with .png, but only the second one is an actual image; you can scroll down the page on the first link and find the “More Details” button) 
![Example link to image stored on Wikimedia](https://upload.wikimedia.org/wikipedia/commons/thumb/d/d3/Albert_Einstein_Head.jpg/180px-Albert_Einstein_Head.jpg)

Example link to image stored on Wikimedia

Making Markdown Tables

Do you dread making tables in markdown? Try using an R data frame and let knitr do the work for you! Here’s an example R code chunk and the web table it produces:

library(knitr)
my.data <- data.frame(
    months = c("Jan", "Feb", "Mar", "Apr", "May", "Jun",
               "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"),
    days_in_month = c(31, 28, 31, 30, 31, 30,
                      31, 31, 30, 31, 30, 31)
    )
knitr::kable(my.data, col.names = c("Months", "Number of Days"))
Months Number of Days
Jan 31
Feb 28
Mar 31
Apr 30
May 31
Jun 30
Jul 31
Aug 31
Sep 30
Oct 31
Nov 30
Dec 31

More on R’s data.frame object can be found here.

Hacking ArcGIS Pro’s Python

https://pro.arcgis.com/en/pro-app/arcpy/get-started/what-is-conda.htm

Converting Images to Animations

Based on Davis’s MODIS EVI example using GIMP and FFmpeg (FFmpeg installed with Imagemagick).

  1. Load spatial data into QGIS

    • Create symbology (save symbology to text file so you can open it and apply it to all your other layers; saves time selecting color ramps and ranges each time)
    • Create layout (save layout in case you need to come back to it)
    • Add map(s), legend(s), and text to layout
    • Update and export maps to images (PNG) for each visualization

  2. Load images into GIMP

    • Open one PNG in GIMP
    • Open as layers the remaining images (order most recent layer at the top)
    • Crop image to selection (use Rectangle Select Tool (shortcut key r) to draw a box around what you want, then use Image → Crop to Selection)
    • Optimize your layers for GIF animation (Filters → Animation → Optimize for GIF)

  3. Save your layers as an animated GIF (Export → GIF)

    • Check [x] As animation
    • Loop forever
    • Delay between frames (e.g., 150 ms)
    • Use delay entered above for all frames

  4. From the command line, convert exported GIF to video format (reference)

    Update the -i input file name and .mp4 output file name

    ffmpeg -f gif -i Untitled.gif -pix_fmt yuv420p -c:v libx264 -movflags +faststart -filter:v crop='floor(in_w/2)*2:floor(in_h/2)*2' Untitled.mp4

Locating Your Google Places

  1. In Google Maps, save locations to starred places
  2. Open Google Takeout
  3. Unselect everything except for Maps (your places)
  4. Choose your delivery method (e.g., email link) and file type (e.g., compressed zip)
  5. Archive will save as GeoJSON
  6. Open GeoJSON in a GIS or other geospatial application

Interpolating

Let’s say you are given two pairs of information (\(x_1, y_1\)) and (\(x_2, y_2\)). These, for example, may represent elevations at two given locations or rainfall values at two given dates. Regardless of what they represent, let’s say you are interested in finding the value \(u\) that lies somewhere between \(y_1\) and \(y_2\), which occurs at a location \(k\) between \(x_1\) and \(x_2\). In table format, it looks something like this:

X Y Description
\(x_1\) \(y_1\) Known \(x,y\) pair just below the value you are looking for
\(k\) \(u\) The pair you want the \(y\)-value for
\(x_2\) \(y_2\) Known \(x,y\) pair just above the value you are looking for

By linear interpolation, the rule of equal ratios applies, which states the following:

\[\frac{k - x_1}{x_2 - x_1} = \frac{u - y_1}{y_2 - y_1}\]

and can be solved for the unknown value, \(u\), represented by the following:

\[u = \frac{\left(y_2 - y_1\right) \left(k - x_1\right)}{x_2 - x_1} + y_1\]

How to Make Multiband Raster in QGIS

Let’s say you have three single-band raster files (e.g., the red, green and blue bands for a color image) and you want to make a raster to visualize the colors. This requires you to merge the three single-band rasters into one multi-band raster. One way to do this in QGIS is to use the Build virtual raster tool found under Raster –> Miscellaneous.

For your input layers, select all your single-band raster files—remember the order in which they are added. All your raster files should have the same resolution, so all options for resolution should result in the same output. Tick the box (or set to True) to place each input file into a separate band. If you raster layers have a NODATA value, be sure to include it here. You can save your virtual raster to file (e.g., GeoTIFF) or just save to memory.