Rounding POSIXct milliseconds in data.table v1.9.2 (ok in 1.8.10)

Question

I have a weird result for my data.table v1.9.2:

 DT timestamp 1: 2013-01-01 17:51:00.707 2: 2013-01-01 17:51:59.996 3: 2013-01-01 17:52:00.059 4: 2013-01-01 17:54:23.901 5: 2013-01-01 17:54:23.914 str(DT) Classes 'data.table' and 'data.frame': 5 obs. of 1 variable: $ timestamp: POSIXct, format: "2013-01-01 17:51:00.707" "2013-01-01 17:51:59.996" "2013-01-01 17:52:00.059" "2013-01-01 17:54:23.901" ... - attr(*, "sorted")= chr "timestamp" - attr(*, ".internal.selfref")=<externalptr> 

When I use the duplicated() function, I get the following result:

 duplicated(DT) [1] FALSE FALSE FALSE FALSE TRUE 

It’s strange to get the 5th line equal to the fourth. It also blocks me from joining tables in R. Is there anything to do with the POSIXct type?

DT on skydrive: DT

Thanks.

Answer 1

Yes, I reproduced your result with v1.9.2.

 library(data.table) DT <- data.table(timestamp=c(as.POSIXct("2013-01-01 17:51:00.707"), as.POSIXct("2013-01-01 17:51:59.996"), as.POSIXct("2013-01-01 17:52:00.059"), as.POSIXct("2013-01-01 17:54:23.901"), as.POSIXct("2013-01-01 17:54:23.914"))) options(digits.secs=3) # usually placed in .Rprofile DT timestamp 1: 2013-01-01 17:51:00.707 2: 2013-01-01 17:51:59.996 3: 2013-01-01 17:52:00.059 4: 2013-01-01 17:54:23.901 5: 2013-01-01 17:54:23.914 duplicated(DT) ## [1] FALSE FALSE FALSE FALSE TRUE 

---

Update v1.9.3 from Matt

In v1.9.2, a rounding change occurred that affected the POSIXct milliseconds. More details here:

Grouping very small numbers (e.g. 1e-28) and 0.0 in data.table v1.8.10 vs v1.9.2

Large integers in data.table. Grouping of results is different in 1.9.2 compared to 1.8.10

So, the workaround available in version 1.0 is:

 > setNumericRounding(1) # default is 2 > duplicated(DT) [1] FALSE FALSE FALSE FALSE FALSE 

I hope you understand why the change was made and agree that we are going in the right direction.

Of course, you do not need to call setNumericRounding() , this is just a workaround.

I registered a new tracker item:

# 5445 numeric rounding should be 0 or 1 automatically for POSIXct