In r4ds/bookclub-dsieur: Data Science in Education Using R Book Club

Walkthrough 4: Longitudinal Analysis With Federal Students With Disabilities Data

Learning objectives:

Longitudinal analysis uses
Importing data options
Processing considerations: what order to take, what to consider for NA's
Importance of visualizing data
Considering next steps for modeling analysis

Slides

r knitr::include_url("https://r4ds.github.io/bookclub-dsieur/R/2021-03-31/Walkthrough-4.html#1")

Meeting Videos

Cohort 1

r knitr::include_url("https://www.youtube.com/embed/6k_F6XGEmqs")

Meeting chat log

00:41:00    edgar zamora:   https://stackoverflow.com/questions/4862178/remove-rows-with-all-or-some-nas-missing-values-in-data-frame
00:41:14    Mike Haugen:    For time series analysis, some of the forecasting functions, e..g exponential smoothing, require a certain approach to dealing with NAs. You can remove NAs for some, for others, you need to impute them
00:42:20    Arami:  Can you explain what "time series analysis" is? Is it any analysis that tracks change over time?
00:43:09    Mike Haugen:    Yes
00:43:48    Mike Haugen:    Like forecasting emergency room presentations based on historical data on emergency room presentations over the last few years
00:44:07    Mike Haugen:    or forecasting course attendance based on historical data.
00:45:18    Rob Lucas:  Glad to know there is some other list-aversion out there!
00:45:27    Mike Haugen:    For R, see Hyndman Forecasting: Principles and Practice: https://otexts.com/fpp3/index.html
00:45:46    Arami:  Thanks!
00:45:47    Ronak Patel:    I also suffer from severe list-aversion.
00:46:33    Morgan Grovenburg:  injuries %>%
  mutate(diag = fct_lump(fct_infreq(diag), n = 5)) %>%
  group_by(diag) %>%
  summarise(n = as.integer(sum(weight)))