Forum Discussion

Anonymous's avatar
Anonymous
Not applicable
5 years ago
Solved

convert to date format

Hi, I am getting below error in date column. could you please help to convert to date format (mm/dd/yyyy)   "power bi DataFormat.Error: We couldn't parse the input provided as a Date value. Detail...
Need Help
  • zzcc's avatar
    zzcc
    5 years ago

    Hi, Anonymous 

     

    column:

    Column = 
    IFERROR(FORMAT(('Table'[Column1]-DATE(1899,12,30))+DATE(1899,12,30),"mm/dd/yyyy"),BLANK())

    measure:

    Measure = 
   IFERROR(FORMAT((MAX('Table'[Column1])DATE(1899,12,30))+DATE(1899,12,30),"mm/dd/yyyy"),BLANK())

    result:

     

     

     

    Zerrick

    Did I answer your question? Mark my post as a solution!

     

zzcc's avatar
zzcc
Frequent Visitor
5 years ago

Hi, Anonymous 

To create a measure like this:

 

Measure = (MAX('Table'[Column1])-DATE(1899,12,30))+DATE(1899,12,30)

 

or create a column:

 

col = ([Column1]-DATE(1899,12,30))+DATE(1899,12,30)

 

result:

 

Zerrick

Did I answer your question? Mark my post as a solution!

 
  • Anonymous's avatar
    Anonymous
    Not applicable
    5 years ago

    Thank you zerick.

     

    But requirement is to convert this serial format(44236, 44239) to MM/DD/YYYY in query editor 

     

    Suren

    • zzcc's avatar
      zzcc
      Frequent Visitor
      5 years ago

      Hi, Anonymous 

      measure

       

      Column = 
var _1=([Column1]-DATE(1899,12,30))+DATE(1899,12,30)
return FORMAT(_1,"mm/dd/yyyy")

       

      column

       

      Measure = 
var _1=(MAX('Table'[Column1])-DATE(1899,12,30))+DATE(1899,12,30)
return FORMAT(_1,"mm/dd/yyyy")

       

       result:

      In power query

       

      let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WMtY3MtU3MjAyVIrViVYyMTEyNgOzgOJGqOKWSrGxAA==", 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}}),
    #"Added Custom" = Table.AddColumn(#"Changed Type", "Custom", each try (Duration.From([Column1]))
otherwise null),
    #"Added Custom1" = Table.AddColumn(#"Added Custom", "Custom.1", each try(Date.From([Column1]))
otherwise null),
    #"Changed Type1" = Table.TransformColumnTypes(#"Added Custom1",{{"Custom", type duration}}),
    #"Added Custom2" = Table.AddColumn(#"Changed Type1", "Custom.2", each #date(1899,12,30)+[Custom]),
    #"Merged Columns" = Table.CombineColumns(Table.TransformColumnTypes(#"Added Custom2", {{"Custom.1", type text}, {"Custom.2", type text}}, "en-US"),{"Custom.1", "Custom.2"},Combiner.CombineTextByDelimiter("", QuoteStyle.None),"Merged"),
    #"Removed Columns1" = Table.RemoveColumns(#"Merged Columns",{"Custom", "Column1"}),
    #"Changed Type2" = Table.TransformColumnTypes(#"Removed Columns1",{{"Merged", type date}})
in
    #"Changed Type2"

       

       

       

      Zerrick

      Did I answer your question? Mark my post as a solution!

      • Anonymous's avatar
        Anonymous
        Not applicable
        5 years ago

        Thank you Zerrick.

        Below expression works for me but encounter errors "Cannot convert value '' of type Text to type Number." when column has "" or "N/A". Could you please help me to fix 

         

        Column2 =
        var _1=([Column1]-DATE(1899,12,30))+DATE(1899,12,30)
        return FORMAT(_1,"mm/dd/yyyy")
  • mahoneypat's avatar
    mahoneypat
    Microsoft Employee
    5 years ago

    You can add a custom column (or adapt a transform column step) with this formula

     

    = try Date.From([Date]) otherwise Date.AddDays(#date(1900,1,1), Number.FromText([Date]))

     

    Pat

     

    • zzcc's avatar
      zzcc
      Frequent Visitor
      5 years ago

      Hi, mahoneypat 

       

      You are absolutely right, obviously I am still not familiar with the PoweQuery language. It definitely saves a lot of steps.
      But what confuses me is that in Excel it seems that the first day is January 1, 1900, and the result I get after converting the numbers to dates in Power BI indicates that the first day should be December 30, 1899, which is a little strange.
      But your formula is definitely correct anyway. Thanks for sharing.

       

      Anonymous  if you want to get result in Power Query, here is the improved M code:

       

      let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WMtY3MtU3MjAyVIrViVYyMTEyNgOzgOJGqOKWYBaYUFCKjQUA", 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}}),
    a = Table.AddColumn(#"Changed Type", "1899-12-30", each try Date.From([Column1]) otherwise Date.AddDays(#date(1899,12,30), Number.FromText([Column1])))//,
    //b = Table.AddColumn(a, "1900-1-1", each try Date.From([Column1]) otherwise Date.AddDays(#date(1900,1,1), Number.FromText([Column1])))
in
    a//b

       

       

      Zerrick

      Did I answer your question? Mark my post as a solution!

      • mahoneypat's avatar
        mahoneypat
        Microsoft Employee
        5 years ago

        Thanks for that correction.  I didn't check the math on the date change.  I checked it with today's date (44409 as integer in Excel), and had to update the formula to this instead.

         

        = try Date.From([Date]) otherwise Date.AddDays(#date(1899,12,31), Number.FromText([Date])-1)

         

        Pat

         

    • Anonymous's avatar
      Anonymous
      Not applicable
      5 years ago

      Thank you Pat. Your expression works well but when date column has some text values(N/A..) it shows error

       

      Suren