how to find number of unique rows for combination of r columns

Sorry, but you ask basic questions.You really need to spend some more time
with an R tutorial or two. This list is not meant to replace your own
learning efforts.

You also do not seem to be reading the docs carefully. Under ?unique, it
links ?duplicated and tells you that it gives indices of duplicated rows of
a data frame. These then can be used by subscripting to remove those rows
from the data frame. Here is a reproducible example:

df <- data.frame(a = 1:3, b = letters[c(1,1,2)], d = LETTERS[c(1,1,2)])
df[-duplicated(df[,2:3]), ]  ## Note the - sign

If you prefer, the "Tidyverse" world has what are purported to be more
user-friendly versions of such data handling functionality that you can use
instead.


Bert

On Fri, Nov 8, 2019 at 7:38 AM Ana Marija <sokovic.anamarija at gmail.com>
wrote:

would you know how would I extract from my original data frame, just
these unique rows?
because this gives me only those 3 columns, and I want all columns
from the original data frame

head(udt)

   chr   pos         gene_id
1 chr1 54490 ENSG00000227232
2 chr1 58814 ENSG00000227232
3 chr1 60351 ENSG00000227232
4 chr1 61920 ENSG00000227232
5 chr1 63671 ENSG00000227232
6 chr1 64931 ENSG00000227232

head(dt)

    chr   pos         gene_id pval_nominal pval_ret       wl      wr
META
1: chr1 54490 ENSG00000227232     0.608495 0.783778 31.62278 21.2838
0.7475480
2: chr1 58814 ENSG00000227232     0.295211 0.897582 31.62278 21.2838
0.6031214
3: chr1 60351 ENSG00000227232     0.439788 0.867959 31.62278 21.2838
0.6907182
4: chr1 61920 ENSG00000227232     0.319528 0.601809 31.62278 21.2838
0.4032200
5: chr1 63671 ENSG00000227232     0.237739 0.988039 31.62278 21.2838
0.7482519
6: chr1 64931 ENSG00000227232     0.276679 0.907037 31.62278 21.2838
0.5974800

On Fri, Nov 8, 2019 at 9:30 AM Ana Marija <sokovic.anamarija at gmail.com>
wrote:

Thank you so much! Converting it to data frame resolved the issue!

On Fri, Nov 8, 2019 at 9:19 AM Gerrit Eichner
<gerrit.eichner at math.uni-giessen.de> wrote:

It seems as if dt is not a (base R) data frame but a
data table. I assume, you will have to transform dt
into a data frame (maybe with as.data.frame) to be
able to apply unique in the suggested way. However,
I am not familiar with data tables. Perhaps somebody
else can provide a more profound guess.

  Regards  --  Gerrit

---------------------------------------------------------------------
Dr. Gerrit Eichner                   Mathematical Institute, Room 212
gerrit.eichner at math.uni-giessen.de   Justus-Liebig-University Giessen
Tel: +49-(0)641-99-32104          Arndtstr. 2, 35392 Giessen, Germany
http://www.uni-giessen.de/eichner
---------------------------------------------------------------------

Am 08.11.2019 um 16:02 schrieb Ana Marija:

I tried it but I got this error:

udt <- unique(dt[c("chr", "pos", "gene_id")])

Error in `[.data.table`(dt, c("chr", "pos", "gene_id")) :
   When i is a data.table (or character vector), the columns to

join by

must be specified using 'on=' argument (see ?data.table), by keying

x

(i.e. sorted, and, marked as sorted, see ?setkey), or by sharing
column names between x and i (i.e., a natural join). Keyed joins

might

have further speed benefits on very large data due to x being sorted
in RAM.

On Fri, Nov 8, 2019 at 8:58 AM Gerrit Eichner
<gerrit.eichner at math.uni-giessen.de> wrote:

Hi, Ana,

doesn't

udt <- unique(dt[c("chr", "pos", "gene_id")])
nrow(udt)

get close to what you want?

   Hth  --  Gerrit

---------------------------------------------------------------------

Dr. Gerrit Eichner                   Mathematical Institute, Room

212

gerrit.eichner at math.uni-giessen.de   Justus-Liebig-University

Giessen

Tel: +49-(0)641-99-32104          Arndtstr. 2, 35392 Giessen,

Germany

http://www.uni-giessen.de/eichner

---------------------------------------------------------------------

Am 08.11.2019 um 15:38 schrieb Ana Marija:

Hello,

I have a data frame like this:

head(dt,20)

       chr    pos         gene_id pval_nominal  pval_ret

 wl      wr

   1: chr1  54490 ENSG00000227232    0.6084950 0.7837780 31.62278

21.2838

   2: chr1  58814 ENSG00000227232    0.2952110 0.8975820 31.62278

21.2838

   3: chr1  60351 ENSG00000227232    0.4397880 0.8679590 31.62278

21.2838

   4: chr1  61920 ENSG00000227232    0.3195280 0.6018090 31.62278

21.2838

   5: chr1  63671 ENSG00000227232    0.2377390 0.9880390 31.62278

21.2838

   6: chr1  64931 ENSG00000227232    0.2766790 0.9070370 31.62278

21.2838

   7: chr1  81587 ENSG00000227232    0.6057930 0.6167630 31.62278

21.2838

   8: chr1 115746 ENSG00000227232    0.4078770 0.7799110 31.62278

21.2838

   9: chr1 135203 ENSG00000227232    0.4078770 0.9299130 31.62278

21.2838

10: chr1 138593 ENSG00000227232    0.8464560 0.5696060 31.62278

21.2838

it is very big,

dim(dt)

[1] 73719122        8

To count number of unique rows for all 3 columns: chr, pos and

gene_id

I could just join those 3 columns and than count. But how would I

find

unique number of rows for these 4 columns without joining them?

Thanks
Ana