Removing Numbers

[chmoon]

I was wondering if anyone knew of a more elgant way to remove numbers from a string. I realized that Power BI does not support regular expressions. I know I could use R or another method to remove, but I am working on training users that will primarily only be using Power BI and Excel for simple data analysis. 

MeterType = SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(FileToBeCleaned[itype],"0",""),"1",""),"2",""),"3",""),"4",""),"5",""),"6",""),"7",""),"8",""),"9","")

Does not work 😞

MeterType = SUBSTITUTE(ToBeCleaned[itype],%[0-9]%,"")

[MarcelBeug]

Thanks for kudos, guys, but - shame on me  - today I was triggered by a question from @MattAllington on MrExcel.com involving Text.Remove, which can be used instead of my creative formula, which can be simply replaced by:

Text.Remove([TextWithDigits],{"0".."9"})

[Anonymous]

Just what I wanted! Thanks!

[rtse]

Hi, I'm new to PowerBI and SQL.  I have used your recommended formula to remove numbers from a text string but PowerBI is giving me an error.  What am I doing wrong ?  thanks

[rtse]

I reinput the command and it worked !!!

[Sean]

@MarcelBeug  Thanks for the Update!

I will actually use this in place of Step 2 here

http://community.powerbi.com/t5/Desktop/split-multiple-column-values-into-multiple-rows/m-p/144081#M...

This Step 2 seems to remove everything after a number - so if a cell starts with a number it will be left empty

= Table.TransformColumns(#"Unpivoted Other Columns", {{"Attribute", each Text.Start(_, Text.PositionOfAny(_,Text.ToList("0123456789"))), type text}})

[MarcelBeug · Accepted solution]

In the Query Editor you can either add a column with formula:

= Text.Combine(List.RemoveItems(Text.ToList([TextWithDigits]),{"0".."9"}))

Or transform the column: first choose Transform - Clean (or some other option) and then adjust the code in the formula bar to:

= Table.TransformColumns(PreviousStep,{{"TextWithDigits", each Text.Combine(List.RemoveItems(Text.ToList(_),{"0".."9"}))}})

the adjustment is the part starting with "each".

[Phil_Seamark]

You can use R to run a regular expression over your column if that helps.

If you give me your preferred Regex pattern I can try and get it going for you.

[Phil_Seamark]

The R column in Power Query might look like this :

Just replace Column1 with your own column.  This will append on a new column stripped of numbers.

This allows you to extend your regular expression to handle other patterns, so could be quite versatile

pattern<-"[A-Z]+"
output <-within(dataset,{RegExCol=sapply(sub('[0-9]+', '', dataset$Column1), function(x) toString(x)) })

[Vvelarde]

@chmoon

Hi, another Option is Query Editor..Use the Replace. Repeating for each number.