Finding all possible combinations for these set conditions

[Arbitas]

ID Value Time
1 0.3 1/3/19 16:00
1 0.2 1/3/19 18:00
1 0.6 1/4/19 12:00
1 0.4 1/4/19 20:00
2 0.1 1/3/19 13:00
2 0.2 1/4/19 19:00
2 0.2 1/5/19 07:00
2 0.4 1/5/19 11:00

I am a beginner in R and am stuck on how to accomplish the following task:

I want to know, for every unique ID, is there an increase of "Value" greater or equal to >=0.3 WITHIN 48 hours.

So essentially, the following checks need to be done

For every ID, if Time2-Time1 <= 48 hrs, check if there is an instance of the value increasing by >=0.3.

The problem is, the value increasing by >=0.3 doesn't always happen right next to each other (for example for ID 2, the value does increase by 0.3 within 48 hours, but it isn't subsequent)

Any help would be appreciated.

[gitdemont]

Hi,

# set min reprex
df = as.data.frame(matrix(c(1,0.3,"1/3/19 16:00",
                            1,0.2,"1/3/19 18:00",
                            1,0.6,"1/4/19 12:00",
                            1,0.4,"1/4/19 20:00",
                            2,0.1,"1/3/19 13:00",
                            2,0.2,"1/4/19 19:00",
                            2,0.2,"1/5/19 07:00",
                            2,0.4,"1/5/19 11:00"), ncol = 3, byrow = TRUE))
colnames(df) = c("ID", "Value", "Time")

# convert to time
df$Time2  = as.POSIXct(df$Time, format = "%d/%m/%Y %H:%M")

# set min diff time to take into account
x = 48
attr(x, "units") = "hours"

by(df, df$ID, FUN = function(d) {
  c(NA, (d$Value[-1] >= 0.3) & (diff(d$Time2) >= x))
})

[HanOostdijk]

The solution by @gitdemont does not satisfy
The problem is, the value increasing by >=0.3 doesn't always happen right next to each other (for example for ID 2, the value does increase by 0.3 within 48 hours, but it isn't subsequent)

library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
library(tibble)
library(lubridate)
#> 
#> Attaching package: 'lubridate'
#> The following objects are masked from 'package:base':
#> 
#>     date, intersect, setdiff, union

df <- tibble::tribble(
  ~ID,~Value,~Time,
1, 0.3, "1/3/19 16:00",
1, 0.2, "1/3/19 18:00",
1, 0.6, "1/4/19 12:00",
1, 0.4, "1/4/19 20:00",
2, 0.1, "1/3/19 13:00",
2, 0.2, "1/4/19 19:00",
2, 0.2, "1/5/19 07:00",
2, 0.4, "1/5/19 11:00"
)

df2 <- df %>%
  mutate(
         nr = row_number(),  
         Time=lubridate::parse_date_time(Time,"%m/%d%y %H:%M")
         )

as.data.frame(df2)
#>   ID Value                Time nr
#> 1  1   0.3 2019-01-03 16:00:00  1
#> 2  1   0.2 2019-01-03 18:00:00  2
#> 3  1   0.6 2019-01-04 12:00:00  3
#> 4  1   0.4 2019-01-04 20:00:00  4
#> 5  2   0.1 2019-01-03 13:00:00  5
#> 6  2   0.2 2019-01-04 19:00:00  6
#> 7  2   0.2 2019-01-05 07:00:00  7
#> 8  2   0.4 2019-01-05 11:00:00  8

df3 <- df2 %>%
  inner_join(df2,by=c('ID'='ID')) %>% 
  mutate(diffTime=lubridate::make_difftime(Time.y - Time.x,units="hours"),
         diffValue = Value.y - Value.x
         ) %>%
  filter((diffTime > 0) & (diffTime <= 48 ) & diffValue >= 0.3)

as.data.frame(df3)
#>   ID Value.x              Time.x nr.x Value.y              Time.y nr.y diffTime
#> 1  1     0.3 2019-01-03 16:00:00    1     0.6 2019-01-04 12:00:00    3 20 hours
#> 2  1     0.2 2019-01-03 18:00:00    2     0.6 2019-01-04 12:00:00    3 18 hours
#> 3  2     0.1 2019-01-03 13:00:00    5     0.4 2019-01-05 11:00:00    8 46 hours
#>   diffValue
#> 1       0.3
#> 2       0.4
#> 3       0.3
Created on 2021-07-26 by the reprex package (v2.0.0)

[gitdemont]

You are right indeed @HanOostdijk, I was totally wrong !
Besides date conversion was also wrong in my solution

Maybe this one is better

# set min reprex
df = as.data.frame(matrix(c(1,0.3,"1/3/19 16:00",
                            1,0.2,"1/3/19 18:00",
                            1,0.6,"1/4/19 12:00",
                            1,0.4,"1/4/19 20:00",
                            2,0.1,"1/3/19 13:00",
                            2,0.2,"1/4/19 19:00",
                            2,0.2,"1/5/19 07:00",
                            2,0.4,"1/5/19 11:00"), ncol = 3, byrow = TRUE))
colnames(df) = c("ID", "Value", "Time")

# convert to time
df$Time2  = as.POSIXct(df$Time, format = "%m/%d/%y %H:%M")

df

by(df, df$ID, FUN = function(d) {
  if(nrow(d) == 1) return(NULL)
  apply(combn(1:nrow(d), 2), 2, FUN = function(i_comb) {
      if(diff(as.numeric(d[i_comb, "Value"])) >= 0.3 & difftime(df[i_comb[2], "Time2"], df[i_comb[1], "Time2"], units = "hours") <= 48)
        return(d[i_comb,])
    })
})

EDIT: sorry for the multiple editions

[gitdemont]

and another one to remove NULL

by(df, df$ID, FUN = function(d) {
  if(nrow(d) == 1) return(NULL)
  Filter(Negate(is.null), apply(combn(1:nrow(d), 2), 2, FUN = function(i_comb) {
      if(diff(as.numeric(d[i_comb, "Value"])) >= 0.3 & difftime(df[i_comb[2], "Time2"], df[i_comb[1], "Time2"], units = "hours") <= 48)
        return(d[i_comb,])
    }))
})

[Arbitas]

Thank you both so much for the help