ous-shiny-demo

This is an example shiny dashboard used for monitoring Ocean Use Survey results from SeaSketch. The data contained in this repo were entirely fabricated for demonstration purposes.

Use as a framework

This app is designed to be both a functional example and a template that can be adapted to monitor other, real SeaSketch surveys. The goal was to generalize as much of the code as possible, allowing users to make minimal changes to apply it to their unique survey. That being said, surveys in SeaSketch can vary widely, and the amount of work required to adapt it to a given survey will also vary.

Adaptation workflow

These are the basic steps required to adapt this framework to your survey

• Clone the repo
• Replace data/responses.csv and data/shapes.json with the respective survey exports from your SeaSketch project
  • Note that the spatial data from SeaSketch will defualt to a different name - change the file name to shapes.json
• Replace data/eez.fgb with a flatgeobuf containing the EEZ or other study area polygon applicable to your project
• Add data/survey_targets.csv with your own targets CSV containing the columns "sector" and "target"
• Edit R/project_variables.R - this script contains all project-specific variables:
  • project is the prefix added to all downloads from the app
  • app_title will be diplayed in the upper left hand corner of the app ui
  • seasketch_url should be the link to your seasketch project page and is accessible via the "Responses Updated:" box on the "Overview" tab
  • sectors is a character vector of ocean use sectors in your project
  • region is the export id of your region question
  • region_list is a character vector of the regions respondents can choose
  • age_groups is a character vector of the age groups respondents can choose from
  • genders is a character vector of the genders respondents can choose from
  • json_columns is a dataframe containing the export ids of any group response questions (which export as JSON-formatted strings), and their respective answer option objects (age_groups and genders in this example) - this excludes region because it is parsed differently from the other group response questions
  • columns_to_remove is a character vector of any columns you don't need in your cleaned data
  • shape_attributes_to_keep is a character vector of shape attributes you want to keep in your cleaned spatial data
• Run R/data_init.R - this will process the data in responses.csv and shapes.json, writing five *.RDS files to data/temp:
  • respondent_info.RDS contains aggregated response data, with one row per survey response (similar to how it's exported from SeaSketch)
  • responses.RDS contains response data with one row per sector response
  • shapes.RDS contains the spatial data joined with other response data in the form of an SF dataframe
  • data_date.RDS contains the date that responses.csv was created
  • temp_data_date.RDS contains the date that the files in data/temp were created (i.e. when R/data_init.R was run)
• Secure app
  • If your app doesn't require authentication, skip to the last bullet in this section - otherwise, follow the steps below
  • Create user database
    • Source auth/create_user_db.R
    • Run create_user_db({initial_user}, {password}, {passphrase}) with the arguments:
      • initial_user: your desired username to access the app
      • password: your desired password
      • passphrase your desired passphrase to decrypt the database
    • This will create auth/users.sqlite
    • This app implementation stores the passphrase in a (gitignored) plaintext file auth/passphrase.txt which is read in global.R - consider whether this is secure enough for your needs
    • Once the app is up and running you can use initial credentials created with create_user_db() to login and any additional users can be added from the shinymanager admin interface accessible via the plus icon in the lower righthand corner of the app
  • Add R/secure_option.R which should assign TRUE to the variable secure - this is intended to be an easy gitignored way to develop locally with secure set to FALSE. Just make sure it's set to TRUE wherever it's deployed. If you don't need your app secured, just set secure to FALSE and call it a day

Updating data

Currently, SeaSketch doesn't have a developer API for requesting survey data. Updates will need to be done by manually adding the latest responses.csv and shapes.json files exported from SeaSketch to the shiny app repo.