IF Column Contains Specific Character followed by String

I have a series of columns. The first row contains dates. Each section of columns ends with a total column that is structured with the year followed by a forward slash symbol, and the month for the particular data set.

I want to be able to write an IF statement where I can single out these total columns, then convert all values in those columns to null (so ultimately these columns can be dynamically filtered out of the data set).

It is important to note that the position of these total columns changes with different data inputs. In the screenshot below, it happens to be in column F34, but if I used, for example, April data, it may be in column F46, or something along those lines.

I'm thinking the formula would be something like the following:

IF [any column] CONTAINS "/*" THEN NULL() ELSE NULL()

The * would represent a wildcard character for any string value, or value that is not a number following the forward slash symbol.

Any help here?

Date Time

Data Investigation

Accepted answers

mceleavey

Hi @lp_hall ,

If you want them removed you can simply do the following:

This isolates the first row, determines if any of the fields are of the shorter length, makes the value null if they are and removes null columns, then uses a specific config on the union to only take a sub-set of all inputs:

Hope this helps,

mceleavey_solution.yxmd

atcodedog05

Hi @lp_hall

Here is how it can be done. And you were right about join 😅. We can use join tool join on position.

Workflow:

Hope this helps 🙂

lp_hall Join First Two Fields (1).yxmd

All comments

apathetichell

sorry just saw- that you want to filter out these columns - easiest way is probably

DYNAMIC RENAME FIRST (FIRST ROW OF DATA)

record id.

transpose. (record id as key column)

formula tool:

if regex_match([name],"\d{4}\/[a-z]{3}") then null() else [value] endif

and then cross-tab back (record id as key column again)

or just filter on !regex_match([name],"\d{4}\/[a-z]{3}" and take the false anchor and then crosstab.

dynamic select formula:

!regex_match([name],"\d{4}\/[a-z]{3}"

should work too.

mceleavey

Hi @lp_hall ,

If you want them removed you can simply do the following:

Hope this helps,

mceleavey_solution.yxmd

mceleavey

@lp_hall ,

Alternatively, you can dynamically rename from the first row of data, then dynamically select those columns that are ten characters in length:

mceleavey_solution2.yxmd

lp_hall

@mceleavey,

Thank you so much! This did exactly what I needed it to do. I appreciate the quick response as well!

@apathetichell,

I appreciate your response and willingness to help as well.

lp_hall

Hey again @mceleavey,

Perhaps I spoke too soon... while your solution does remove the row 1 header, the column itself still exists after your workflow runs. I want the entire column to be filtered out/removed, not just the header.

Can you modify your workflow to have this additional step added please? Again thanks for all the help so far!

apathetichell

Here's how my methodology would work...

note (stole the input file from @mceleavey out of laziness)

regex transpose solution.yxmd

mceleavey

@apathetichell

“We often miss opportunity because it's dressed in overalls and looks like work”
― Thomas A. Edison

Quick Links

This months top contributors

atcodedog05 19458

Qiu 15866

binu_acs 15708

MarqueeCrew 13708

apathetichell 13703