CODEWITHSUNDEEP
rdataframeroundingna
user3385769
user3385769

Reputation: 161

Round data all data in dataframe with NA's

I have a data frame which has NA's in the data. I can easily reformat a whole column to 3 decimal places disregarding the NA's. However, I can't find a way to do the entire dataframe. is there no easier way to do this than separately operating on the columns? I didn't have any luck getting apply to work either. 

> ndf3
                    a                  b                 c                   d                  e
1   0.144346473029046 0.0945535269709544 0.127334024896266   0.125881659751037  0.108239419087137
2              0.1376             0.1489            0.0949             0.11784             0.1372
3    1.48794219374094   1.53595988141984  1.21977361926472    1.15852039195536  0.707774430272806
4    3.96992500677986   3.90772031563767  3.61382453467775    3.39967072835538   4.30292753254034
5   0.174889714039075 -0.419130501555829 0.113585656836338 -0.0307388258550179 -0.309314899645693
6   0.135682205983122  0.189426931757576 0.158335341568249   0.116675569765867 0.0645715900731325
252  37.8736196377026   22.0640317131964  33.4935080490794    33.2572127571306   29.0085133196937
8   0.335796353933727  0.179593968145971 0.484609677920302                  NA                 NA
9   0.788381742738589  0.742738589211618 0.767634854771784   0.813278008298755                  1
10   1.42703970549072   1.33470191247724  1.25827986596585    1.32419222712559                  1

> class(ndf3)
[1] "data.frame"
> typeof(ndf3)
[1] "list"
> c <- format(round(as.numeric(as.character(ndf3)),3), nsmall = 3, na.encode = TRUE)
Warning message:
In format(round(as.numeric(as.character(ndf3)), 3), nsmall = 3,  :
  NAs introduced by coercion
> c
[1] "NA" "NA" "NA" "NA" "NA"


> is.character(ndf3)
[1] FALSE
> is.character(ndf3[[1]])
[1] TRUE
> round(as.numeric(ndf3[[1]]),3)
 [1]  0.144  0.138  1.488  3.970  0.175  0.136 37.874  0.336  0.788  1.427

Upvotes: 0

Views: 2173

Answers (2)

vpipkt
vpipkt

Reputation: 1707

If every column in the data frame is numeric: as.data.frame(round(as.matrix(ndf3),3))

Upvotes: -1

arvi1000
arvi1000

Reputation: 9582

If you just want to round to 3 digits:

ndf3 <- data.frame(lapply(ndf3, round, 3))

Edit: the above method lets you apply any operation to a whole data.frame. E.g., to coerce to numeric and then round to 3 digits:

lapply(ndf3, function(x) round(as.numeric(x), 3))

Or if you prefer not to use an anonymous function, you can define the function in advance. This is equivalent to the above:

forced_numeric_round3 <- function(x) round(as.numeric(x), 3)
lapply(ndf3, forced_numeric_round3)

Upvotes: 1

Related Questions