Loops to close

  • Assignment 3: variants of Getis-Ord cluster analysis
  • Managing data for group projects
    • challenges with Git?
    • Census & other government data can go in the public folder
    • Save documentation about metadata (e.g. dictionary of column names)
    • Prep the simplest, most anonymized RDS data files possible for the app
    • if you must map points, consider randomizing the location somewhat
  • How Prof Holler geocoded profile data
  • Is the geography of New York City a problem?
  • Presentations: 6 minute story with Question, app driver & narrator(s). Focus on your findings and to some extent your achievements, not your struggles
  • qdapRegex package has some useful text cleaning functions for Twitter text, including rm_twitter_url, rm_url, rm_tag, and rm_hash. It also has some functions for address components.
  • The qdapRegex functions above will work best if you convert the tweet text to lower-case first, with the tolower() function.

Possible Project Improvements

  • You may use a batch of geocoded profile locations! Just join this table to the Tweets by the location column and integrate these coordinates into your analysis as the 3rd best option for mapping a Tweet.
  • You may download other datasets from the Census or American Community Survey to include more detailed socio-economic information. Here is a video tutorial from GEOG 261. You don’t need any of the QGIS stuff… the first few videos are a tour of Census data and techniques to download it using tidycensus.
  • You may download data on storm impacts from NOAA’s database. Bulk downloads are available by year, and choose the “details” data series (as opposed to just “locations”) NOAA storm events
  • You may download data on the counties included in Federal Disaster Declarations from FEMA. Hints: filter the CSV file by year, by incidentType, and having “Ida” in the declarationTitle. To join to counties, notice that the placeCode is a 5-digit code composed of the state (2 digits) and county (3 digits), which should be equivalent to county GEOID’s. This data does not indicate how much damage, other than whether there was enough to cross the threshold to provide individual assistance (iaProgramDeclared).
  • Clean tweet text with general expressions

Examples of Great Stories

Best Practices

  • Be clear about the “universe” of your data. What population does the data represent? Is it a complete enumeration or a sample?
  • Use consistent symbology, colors, fonts, and classification schemes
  • Keep interactive graph axes constant
  • Cite and link data sources and any significant sources of code

Thanks to Librarian Julia Deen for these resources:

  • https://medium.com/agoda-engineering/10-common-data-visualization-mistakes-and-how-to-avoid-them-e3896fe8e104
  • Wong, B. Gestalt principles (Part 1). Nat Methods 7, 863 (2010). https://doi.org/10.1038/nmeth1110-863
  • Wong, B. Gestalt principles (Part 2). Nat Methods 7, 941 (2010). https://doi.org/10.1038/nmeth1210-941
  • Wong, B. Points of view: Points of review (part 2). Nat Methods 8, 189 (2011). https://doi.org/10.1038/nmeth0311-189Wong, B. Color coding. Nat Methods 7, 573 (2010). https://doi.org/10.1038/nmeth0810-573
  • Krzywinski, M., Wong, B. Plotting symbols. Nat Methods 10, 451 (2013). https://doi.org/10.1038/nmeth.2490
  • Gehlenborg, N., Wong, B. Mapping quantitative data to color. Nat Methods 9, 769 (2012). https://doi.org/10.1038/nmeth.2134
  • Gehlenborg, N., Wong, B. Heat maps. Nat Methods 9, 213 (2012). https://doi.org/10.1038/nmeth.1902

Citation

  • Software: list the R version and the packages used for your work
  • Data: Who created it? Where is it available (link, or ? When did you acquire it?
  • Literature: Authors, year, and DOI link (e.g. https://doi.org/…)

List this citational information both in the app and on the project’s README.md file.