Classification

[nevrome]

I just came across this interesting research compendium and scrolled a bit through the code. I believe a construct like this could be replaced with less verbose code:

knossoscemeteries/R/Samples.R

Lines 52 to 63 in bdfb456

	aorist_melt$period<-ifelse(aorist_melt$variable<=1100 & aorist_melt$variable>=980,"SM",
	ifelse(aorist_melt$variable<=970 & aorist_melt$variable>=930,"EPG",
	ifelse(aorist_melt$variable<=920 & aorist_melt$variable>=880,"MPG",
	ifelse(aorist_melt$variable<=870 & aorist_melt$variable>=850,"LPG",
	ifelse(aorist_melt$variable<=840 & aorist_melt$variable>=820,"PGB",
	ifelse(aorist_melt$variable<=810 & aorist_melt$variable>=800,"EG",
	ifelse(aorist_melt$variable<=790 & aorist_melt$variable>=750,"MG",
	ifelse(aorist_melt$variable<=740 & aorist_melt$variable>=720,"LG",
	ifelse(aorist_melt$variable<=710 & aorist_melt$variable>=680,"EO",
	ifelse(aorist_melt$variable<=670 & aorist_melt$variable>=640,"LO",
	ifelse(aorist_melt$variable<=630 & aorist_melt$variable>=600,"A",""
	)))))))))))

For this specific numeric-to-label transformation one can use cut in base R or - more flexible - dplyr::case_when. Cut returns an ordered factor, which could actually be useful here. For example:

x <- iris$Petal.Width

iris$PW_class1 <- base::cut(
  x,
  breaks = c(0, 1, 2, 3),
  labels = c("small", "medium", "large")
)

iris$PW_class2 <- dplyr::case_when(
  x > 0 & x <= 1 ~ "small",
  x > 1 & x <= 2 ~ "medium",
  x > 2 & x <= 3 ~ "large",
  TRUE ~ "" # if you really need to cover this case
)

knitr::kable(tail(iris))

	Sepal.Length	Sepal.Width	Petal.Length	Petal.Width	Species	PW_class1	PW_class2
145	6.7	3.3	5.7	2.5	virginica	large	large
146	6.7	3.0	5.2	2.3	virginica	large	large
147	6.3	2.5	5.0	1.9	virginica	medium	medium
148	6.5	3.0	5.2	2.0	virginica	medium	medium
149	6.2	3.4	5.4	2.3	virginica	large	large
150	5.9	3.0	5.1	1.8	virginica	medium	medium

Sorry for crashing in here with my unsolicited advice. I just thought this might be interesting to you 👍