Forum Discussion

dkernen's avatar
dkernen
Resolver II
5 years ago
Solved

Converting text (Replacing values) within multiple columns in Power Query

Good morning!  I have a simple file I need to massage and I would like to learn how to do it more efficiently.  I have four columns that have values that all begin with "Yes" (Yes please enter by who...
Need Help
  • v-yingjl's avatar
    v-yingjl
    5 years ago

    Hi dkernen ,

    You can try this query:

    let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WclTSUYpMLVbQKMhJTSxOVUjNK0ktUkiqVCjPyM/VtMIqW56RmZyhUJBRWZyZnJmYh6mqoiAnMROfeKxOtJITUBaCcOkFK3MeGAfGAgA=", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Category = _t, UMD = _t, MD = _t, BDEarlyConv = _t, DCDPronIssues = _t]),
    #"Changed Type" = Table.TransformColumnTypes(Source,{{"Category", type text}, {"UMD", type text}, {"MD", type text}, {"BDEarlyConv", type text}, {"DCDPronIssues", type text}}),
    #"AllReplace" = 
        [
            #"Yes (please enter by whom):"= "Yes",
            #"Yes (please enter which physician):" = "Yes",
            #"Yes (please explain):" = "Yes"
        ],
    #"Replace Value" = 
        Table.TransformColumns(
            #"Changed Type",
                {
                    {"UMD", each Record.FieldOrDefault(#"AllReplace",_,_),type text},
                    {"MD", each Record.FieldOrDefault(#"AllReplace",_,_),type text},
                    {"BDEarlyConv", each Record.FieldOrDefault(#"AllReplace",_,_),type text},
                    {"DCDPronIssues", each Record.FieldOrDefault(#"AllReplace",_,_),type text}
                }
        )
in
    #"Replace Value"

     

    Best Regards,
    Community Support Team _ Yingjie Li
    If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

dkernen's avatar
dkernen
Resolver II
5 years ago

Greg - thank you for you swift reply!  Definitely I could strip out anything after the first space (if a space exists).  Are you suggesting that  I create new dummy columns, then delete the ones I don't need, and then rename the new one?  I was perhaps hoping I could do it all in one step.  Is that not possible?

 

Greg_Deckler  

Greg_Deckler's avatar
Greg_Deckler
Community Champion
5 years ago

dkernen 1 Step:

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WqkwtVkgsTklLLFaK1YFyk3ISFRKBZIpCamJKIkI8JRskkZMGlIQjpdhYAA==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Column1 = _t]),
    #"Changed Type" = Table.TransformColumnTypes(Source,{{"Column1", type text}}),
    #"Split Column by Delimiter" = Table.SplitColumn(#"Changed Type", "Column1", Splitter.SplitTextByEachDelimiter({" "}, QuoteStyle.Csv, false), {"Column1.1"})
in
    #"Split Column by Delimiter"