Correct PlotFertilityRatesStats() to use new rates format #83
Conversation
| @@ -43,6 +36,7 @@ | |||
| }) | |||
|
|
|||
| df <- data.table::rbindlist(l) | |||
| df <- subset(df, !startsWith(df$Label, "NA")) | |||
There was a problem hiding this comment.
Do we purposely leave the label blank for zero-fertility group from the PopValues ? it'd be nice to show these zero groups as well
There was a problem hiding this comment.
Ah ... good question. Several thoughts:
- In the special case of fertility, it's mostly accepted that fertility is zero before a certain age and after a certain ago, so plotting those regions is probably not informative. (If the fertility rate of 80-year-old women isn't exactly zero, it's so low as not to affect the population predictions.)
- The NAs happen when a label isn't provided, so the simple workaround for missing rate bands is to provide a label. However, this leads to the next problem ...
- The log of zero is -Inf, so these regions won't show up in the log plot anyway. And elsewhere in the code we assume that zero-rate stuff is intended to stay zero, so the non-log plot will be a perfect straight line.
I'm open to suggestions for a different approach. If you have a suggestion, post a feature request and we can discuss.
| @@ -58,7 +52,9 @@ | |||
| #' Plot Fertility Rates Statistics | |||
| #' | |||
| #' @param results Results list (as returned by \code{RunExperiments()}) | |||
| #' @param se Whether to show standard error intervals, or projection intervals | |||
| #' @param se Whether to show standard error or standard deviation confidence intervals | |||
| #' @param type Plot type (options = "lines", "ribbon", "boxplot") | |||
There was a problem hiding this comment.
the boxplot does not seem to use standard error option, suggest document this
| if (log){ | ||
| g <- ggplot(df, aes( | ||
| x = Year, | ||
| y = LogRate, |
There was a problem hiding this comment.
can this work with y = ifelse(log, LogRate, Rate)?
There was a problem hiding this comment.
I don't think so. As you know, in ggplot column references don't have quotation marks around them, and that implies some tricky parse/evaluate logic inside ggplot. I'm concerned that if we took the ifelse approach, ggplot would go looking for a column called "ifelse(log, LogRate, Rate)". I haven't confirmed this.
Fixes #79