Check break and repeat condition for every 10 consecutive rows-CodePudding

Data

dput

structure(list(DIp = c(12L, 25L, 14L, 17L, 23L, 25L, 34L, 31L, 
27L, 25L, 15L, 16L, 19L, 21L), DIn = c(11L, 25L, 23L, 28L, 17L, 
16L, 21L, 19L, 17L, 20L, 16L, 17L, 21L, 21L)), row.names = c(NA, 
-14L), class = "data.frame")

I want my code to do the following -

If DIp > DIn for at least 10 consecutive rows - mark all the following as Uptrend. Else, mark as Undetermined.

I am trying to do this but I don't know how to break the loop for every 10 consecutive rows- i.e if DIp[1:10] > DIn[1:10], then, return new column 'MC' with 'Uptrend' for all rows where the condition is met.

CodePudding user response：

I think you need a rolling operation. There are lot of packages in R which help you perform such calculation.

Here's one with zoo's rollaplly -

library(dplyr)
library(zoo)

df <- df %>% mutate(result = rollapplyr(DIp > DIn, 10, all, fill = FALSE))

result would have TRUE/FALSE values. If there are at least 10 consective rows where DIp > DIn then it should return TRUE or else FALSE.

If you want output as text "Uptrend" and "Undetermined" you may use an ifelse.

df <- df %>% 
       mutate(result = ifelse(rollapplyr(DIp > DIn, 10, all, fill = FALSE), 
                              "Uptrend", "Undetermined"))

CodePudding user response：

You can also do this in base R

rleid <- rle(df$DIp > df$DIn)
rleid$values <- rleid$lengths > 9L & rleid$values
df$MC <- c("Undetermined", "Uptrend")[inverse.rle(rleid)   1L]

rle gives you the number of consecutive TRUE/FALSE. TRUE means DIp > DIn. Then we recode the rle values only if the number of consecutive TRUEs is greater than 10. Last, we inverse the rle to get back the vector and map TRUE/FALSE to "Uptrend"/"Undetermined".

Output

> df
   DIp DIn           MC
1   12  11 Undetermined
2   25  25 Undetermined
3   14  23 Undetermined
4   17  28 Undetermined
5   23  17 Undetermined
6   25  16 Undetermined
7   34  21 Undetermined
8   31  19 Undetermined
9   27  17 Undetermined
10  25  20 Undetermined
11  15  16 Undetermined
12  16  17 Undetermined
13  19  21 Undetermined
14  21  21 Undetermined

An example of rle:

> rle(c(T,T,T,F,F))
Run Length Encoding
  lengths: int [1:2] 3 2
  values : logi [1:2] TRUE FALSE
> inverse.rle(rle(c(T,T,T,F,F)))
[1]  TRUE  TRUE  TRUE FALSE FALSE