Split up dataframe into list of dataframes based on breaks in one column

Question

I have a dataframe with a list of consecutive measurements of temperature over a period of time. I would like to split these up into separate dataframes based on when the temperature is consecutively above zero. For example:

Date	Time	Temp
2022-05-20	4:16	32
2022-05-20	4:17	16
2022-05-20	4:18	0
2022-05-20	4:19	7
2022-05-20	4:34	75
2022-05-20	4:39	5
2022-05-20	4:48	0
2022-05-20	4:49	0
2022-05-20	4:59	83

The desired output would be:

Date	Time	Temp
2022-05-20	4:16	32
2022-05-20	4:17	16

and

Date	Time	Temp
2022-05-20	4:19	7
2022-05-20	4:34	75
2022-05-20	4:39	5

and

Date	Time	Temp
2022-05-20	4:59	83

Answer 1

A base R approach would be to use split with cumsum:

split(df[df$Temp != 0, ], 
      cumsum(df$Temp == 0)[df$Temp != 0])

Or a much more readable format suggested by @thelatemail

sel <- df$Temp == 0

split(df[!sel,], cumsum(sel)[!sel])

Output for both:

# $`0`
#   Date Time Temp
# 1 2022-05-20 4:16   32
# 2 2022-05-20 4:17   16
# 
# $`1`
#   Date Time Temp
# 4 2022-05-20 4:19    7
# 5 2022-05-20 4:34   75
# 6 2022-05-20 4:39    5
# 
# $`3`
#   Date Time Temp
# 9 2022-05-20 4:59   83

(Using these data)

df <- read.table(text = "Date   Time    Temp
2022-05-20  4:16    32
2022-05-20  4:17    16
2022-05-20  4:18    0
2022-05-20  4:19    7
2022-05-20  4:34    75
2022-05-20  4:39    5
2022-05-20  4:48    0
2022-05-20  4:49    0
2022-05-20  4:59    83", header = TRUE)