Day 5: Task & Activities

Functions, sourcing scripts, working on your R package

1. Setup

  • Create a new repo on GitHub called eds221-day5-activities
  • Clone to create a version-controlled R Project
  • Create subfolders: docs, src, data, figs

2. Make a function, source in a Quarto doc

Citation

NCDENR Stormwater BMP Manual

One established way to calculate the volume of stormwater expected for a watershed (necessary to design best management practices & systems) is the Simple Method, which involves two steps. First, the runoff coefficient \(R_v\) (storm runoff / storm rainfall) is calculated from: \[R_v = 0.05 + 0.9 * I_A\]

Where \(R_v\) is the runoff coefficient (unitless), and \(I_A\) is the fraction of the watershed that is considered “impervious” (unitless). The volume of stormwater that needs to be handled, \(V\) in cubic feet, is then calculated by: \[V=3630 * R_D * R_v * A\] where \(R_D\) is the “design storm rainfall depth” in inches, usually set to 1.0 or 1.5, \(R_v\) is the runoff coefficient calculated above, and \(A\) is the watershed area in acres.

YOUR TASK:

  • Create a new R script in src, saved as storm_runoff.R

  • In the script, create a function called predict_runoff that estimates the storm runoff volume using inputs for the impervious fraction and watershed area (you can use a constant value of 1 for \(R_D\) here). In other words, your function should only require two arguments

  • Add documentation to your function using Roxygen comments for practice

  • Try out your function in the Console to ensure that it works

  • Create a new R Markdown document in docs, saved as runoff_volumes.Rmd

  • Attach the tidyverse and here packages

  • Source your storm_runoff.R script so you are able to use the predict_runoff function in your .Rmd

  • In a code chunk in your runoff_volumes.Rmd, use your predict_runoff function to estimate stormwater volume for a watershed of 182 acres, over a range of estimates for the impervious fraction (from 0.6 to 0.8, by increments of 0.01). Note: you do not need to write a for loop here.

  • Bind your sequence of impervious fractions together with the resulting runoff volume calculated into a data frame

  • Create a ggplot graph that has both dots and connecting lines (i.e., you’ll layer geom_point() and geom_line()). Update axis labels. Export a png of your graph to the figs folder using ggsave.

3. Wild data

For your next task, you will work with the us_tilapia_imports.csv. It exists in your eds221-day5-comp project - copy that file into your data folder for your day 5 activities project. The data are from the USDA Economic Research Service, and contain annual total volumes (in thousands of pounds) of tilapia imports to the United States from different countries.

You can decide if you want to do this all in separate steps, or piped together in sequence, or some combination. Make sure if you pipe things together, you check the output at every step.

  • Create a new .Rmd in your docs folder called us_tilapia_imports.Rmd

  • Attach the tidyverse, here and janitor packages

  • Read in the data as us_tilapia_imports

  • Explore the data. What are the classes of the columns? Remember some tools we’ve used: summary, names, dim, skim, etc.

  • Use pivot_longer() to reshape the data into long format (currently, the variable “year” is spread across multiple columns). Remember to store the output so you will be able to use the reshaped data this creates.

  • Check the class of the year column. What is it, and why do you think that’s the case? Then, coerce the year column to numeric (e.g. using mutate() and as.numeric() in combination)

  • Use dplyr::group_by() %>% summarize() to find the total US tilapia imports by year, store as yearly_tilapia_tot

  • Create a ggplot line graph of total US tilapia imports for all years in yearly_tilapia_tot. Update axis labels (include units as necessary), then export your graph as a .png to figs.

  • Create a subset that only retains imports from Ecuador, Honduras, Costa Rica, and Mexico (you decide what to name this)

  • Create a ggplot graph of total US tilapia imports over time, for those four countries in the subset you created above, separated by country. Update axis labels, add a title, customize your color scheme, update the theme. Export a .jpg of your graph to figs.

4. Add to your R package

For your final task, reopen your R package you started today, and add two new functions. They can do whatever you want as long as:

  • They have at least two required arguments
  • You add documentation for each function with a Roxygen skeleton

Once you’ve added your functions, make sure to devtools::document(), install and restart, and check to make sure your functions are working, and that you can see your documentation. Then:

  • Push your changes back to your repo on GitHub
  • Share the link with someone (they’ll need to re-install your package from GitHub using install_github("username/reponame"))

End Activity Session (Day 5)