R fill data.frame column wise left to right

Question

i guess, there is as solution in the deep of www, but i can't found it. So I ask here with the hope that it will also help others.

I've a data.frame with multiple columns of the same type of columns. Some are empty some not.

a <- c(sku = "12345678", Image.1 = "name1.jpg", Image.2 = "name2.jpg", Image.3 = "", Image.4 = "name4.jpg", Image.5 = "", Image.6 = "", Image.5 = "name7.jpg")

        sku     Image.1     Image.2     Image.3     Image.4     Image.5     Image.6     Image.5 
 "12345678" "name1.jpg" "name2.jpg"          "" "name4.jpg"          ""          "" "name7.jpg" 

Now I want to move the column values to the left so that all empty columns are at the end of the line. Expacted result:

       sku     Image.1     Image.2     Image.3     Image.4     Image.5     Image.6     Image.5 
 "12345678" "name1.jpg" "name2.jpg" "name4.jpg" "name7.jpg"          ""          ""          "" 

Many thanks for your help

Answer 1

You can use nchar to get the number of characters and then subset one using those with more than 0 and than those with 0 characters.

i <- nchar(a)==0
c(a[!i], a[i])
#        sku     Image.1     Image.2     Image.4     Image.5     Image.3     Image.5     Image.6 
# "12345678" "name1.jpg" "name2.jpg" "name4.jpg" "name7.jpg"          ""          ""          "" 

In case the names should stay where they are use setNames :

setNames(c(a[!i], a[i]), names(a))
#        sku     Image.1     Image.2     Image.3     Image.4     Image.5     Image.6     Image.5 
# "12345678" "name1.jpg" "name2.jpg" "name4.jpg" "name7.jpg"          ""          ""          "" 

or overwrite the data of a :

a[] <- c(a[!i], a[i])
#        sku     Image.1     Image.2     Image.3     Image.4     Image.5     Image.6     Image.5 
# "12345678" "name1.jpg" "name2.jpg" "name4.jpg" "name7.jpg"          ""          ""          "" 

Answer 2

You can use order :

a[] <- a[order(a == '')]
a
#        sku     Image.1     Image.2     Image.3     Image.4 
# "12345678" "name1.jpg" "name2.jpg" "name4.jpg" "name7.jpg" 

#    Image.5     Image.6     Image.5 
#         ""          ""          ""