Hi Tobias,
The "Compute" tool does output just one column of data. The
parenthesis are problematic to get rid of, so the starting and
ending values would need to be disposable. Then you could break up
this column with the tool "Convert delimiters to TAB" (commas and
spaces can be converted) followed by "Cut" to just pull out the ones
that you want (watch out for double tabs - or remove them if really
wanted with the tool "
Condense consecutive characters"). And you are correct, using a
workflow to add the columns iteratively is another option.
The good thing about a workflow is that once created, you can run it
as if it was a single "tool", placing it in the tool menu and hiding
intermediate output datasets. Either of these could become a
workflow.
There is the tool "Arithmetic Operations on tables" on the Test
server (https://test.galaxyproject.org/), but this would involve
creating a new separate file that would be about as much work as the
the others.
I didn't see any obvious solutions in the Tool Shed for use in a
local instance (this type of simple operation should be fine on a
basic laptop install), but this is something you could also review.
http://usegalaxy.org/toolshed
Hopefully one of these options will work out for you, even if not a
single step,
Jen
Galaxy team
On 11/18/13 9:59 PM, Tobias Hohenauer
wrote:
Hello,
I am looking for the right way to do a computation using "text
manipulation, compute an expression on every row".
I have a table consisting of 20 columns and about 15.000 rows.
Column 1 is my untreated or control and I would like to normalize
every other column to this control by simple division.
As a result I would like to obtain a table in which column 1 is
set to a value of 1 for every row, and the corresponding values in
the other columns resulting from column x / column 1.
I tried something like c2/c1,c3/c1,c4/c1... on a small testfile
but that results in all normalized values being put into one
column in brackets rather than being put into individual columns.
What would be the right parameters?
I also thought about creating a workflow consisting of successive
divisions of the columns but that feels rather complicated. Is
there an easier way?
Any help would be much appreciated!
Best,
Tobias
--
Jennifer Hillman-Jackson
http://galaxyproject.org