ISNULL fUNCTION

'REPLACE FN ON DISPOSITION COLUMN'

THANKS

SORRY i MEANT ISNULL FUNCTION

You can use case also like this

 declare @cols nvarchar(2000)
 select  @cols = stuff(( select distinct top 100 percent '],[' + convert(varchar,t2.Calldate,105)
                           from    [Source table] as t2
                          order by '],[' + convert(varchar,t2.Calldate,105)
                            for xml path('')), 1, 2, '') + ']'
declare @query nvarchar(4000)
set @query = 'select [Mobile no],' + @cols +
               'from (select [Mobile no],convert(varchar,CallDate,105) CallDate, Case When Disposition is NULL then '-' Else Disposition End as Disposition  from [Source table]) t1
               pivot (max(Disposition) for Calldate in (' + @cols + ')) as p'
exec (@query)

Using ISNULL

 declare @cols nvarchar(2000)
 select  @cols = stuff(( select distinct top 100 percent '],[' + convert(varchar,t2.Calldate,105)
                           from    [Source table] as t2
                          order by '],[' + convert(varchar,t2.Calldate,105)
                            for xml path('')), 1, 2, '') + ']'
declare @query nvarchar(4000)
set @query = 'select [Mobile no],' + @cols +
               'from (select [Mobile no],convert(varchar,CallDate,105) CallDate, ISNULL(Disposition , '-') AS Disposition   from [Source table]) t1
               pivot (max(Disposition) for Calldate in (' + @cols + ')) as p'
exec (@query)

Msg 8117, Level 16, State 1, Line 9
Operand data type nvarchar is invalid for subtract operator.

Pratima,

It is not necessary that I have to go for '-' sign, any other character will do.
Hope am clear

@ Pratima

I tried with ISNULL Function, throwing me error.


Msg 8117, Level 16, State 1, Line 8
Operand data type nvarchar is invalid for subtract operator.

try this

declare @cols nvarchar(2000)
 select  @cols = stuff(( select distinct top 100 percent '],[' + convert(varchar,t2.Calldate,105)
                           from    [Source table] as t2
                          order by '],[' + convert(varchar,t2.Calldate,105)
                            for xml path('')), 1, 2, '') + ']'
declare @query nvarchar(4000)
set @query = 'select [Mobile no],' + @cols +
               'from (select [Mobile no],convert(varchar,CallDate,105) CallDate, ISNULL(Disposition ,
"-") AS Disposition   from [Source table]) t1
               pivot (max(Disposition) for Calldate in (' + @cols + ')) as p'
exec (@query)

No Luck :(
Msg 207, Level 16, State 1, Line 2
Invalid column name '-'.

what about:

declare @cols nvarchar(2000)
 select  @cols = stuff(( select distinct top 100 percent '],[' + convert(varchar,t2.Calldate,105)
                           from    [Source table] as t2
                          order by '],[' + convert(varchar,t2.Calldate,105)
                            for xml path('')), 1, 2, '') + ']'
declare @query nvarchar(4000)
set @query = 'select [Mobile no],' + @cols +
               'from (select [Mobile no],convert(varchar,CallDate,105) CallDate, Case When Disposition is NULL then ''-'' Else Disposition End as Disposition  from [Source table]) t1
               pivot (max(Disposition) for Calldate in (' + @cols + ')) as p'
exec (@query)

Select allOpen in new window

@Thanks Angel for your suggestion.

Unfortunately there is no change I still see the null values.

try this

declare @cols nvarchar(2000)
 select  @cols = stuff(( select distinct top 100 percent '],[' + convert(varchar,t2.Calldate,105)
                           from    [Source table] as t2
                          order by '],[' + convert(varchar,t2.Calldate,105)
                            for xml path('')), 1, 2, '') + ']'
declare @query nvarchar(4000)
set @query = 'select [Mobile no],' + @cols +
               'from (select [Mobile no],convert(varchar,CallDate,105) CallDate, ISNULL(Disposition ,
'''-''') AS Disposition   from [Source table]) t1
               pivot (max(Disposition) for Calldate in (' + @cols + ')) as p'
exec (@query)

Msg 8117, Level 16, State 1, Line 9
Operand data type nvarchar is invalid for subtract operator.

what about:

declare @cols nvarchar(2000)
 select  @cols = stuff(( select distinct top 100 percent '],[' + convert(varchar,t2.Calldate,105)
                           from    [Source table] as t2
                          order by '],[' + convert(varchar,t2.Calldate,105)
                            for xml path('')), 1, 2, '') + ']'
declare @query nvarchar(4000)
set @query = 'select [Mobile no],' + @cols +
               'from (select [Mobile no],convert(varchar,CallDate,105) CallDate, Case When Disposition is NULL then ''-'' Else Disposition End as Disposition  from [Source table]) t1
               pivot (max(convert(varchar,CallDate,105) CallDate, Case When Disposition is NULL then ''-'' Else Disposition End) for Calldate in (' + @cols + ')) as p'
exec (@query)

Select allOpen in new window

Msg 156, Level 15, State 1, Line 2
Incorrect syntax near the keyword 'convert'.

well, actually: is the column value NULL or empty string?
please

@Angel,

Please find the attachment of my Source table content.
Hope you get an idea.
Thnk you
Source-Table.xlsx

The issue is you are pivoting, IMHO, so the NULLs are when a particular row has a column after pivot for which it didn't actually have a row/record of data; therefore, wrapping the original query with ISNULL does nothing as you are finding.

To fix, you need to generate a second listing of volume that has definitions like:
'ISNULL('+col1+',''-'') as '+col1

Not on a full keyboard, but hopefully you get the drift.

Volume = columns
I.e., if the second variable/listing of column names is @colsSel then you would just replace @cols in the select portion of the query with @colsSel. The instance of @cols in the pivot clause would stay the same.

here is my testing code.
I see that the isnull() has to go like this:

create table  [Source table] ( [Mobile No] varchar(100),	calldate datetime, 	Disposition	varchar(10), HUB varchar(10) )
go
set dateformat dmy
go

insert into [Source table] values('423432657','12/09/2010 00:00','PTP','KK')
insert into [Source table] values('12345823','13/09/2010 00:00','PTP','KK')
insert into [Source table] values('2395586','14/09/2010 00:00','PTP','KK')
insert into [Source table] values('23902334','15/09/2010 00:00','PTP','KK')
insert into [Source table] values('34853433','16/09/2010 00:00','PTP','KK')
insert into [Source table] values('2380334','17/09/2010 00:00','PTP','KK')
insert into [Source table] values('32323854','18/09/2010 00:00','PTP','KK')
insert into [Source table] values('23823844','19/09/2010 00:00','PTP','KK')
insert into [Source table] values('2380232','20/09/2010 00:00','NB','KK')
insert into [Source table] values('83235671','21/09/2010 00:00','NB','KK')
insert into [Source table] values('84235671','22/09/2010 00:00','NB','AP')
go

declare @cols1 nvarchar(max)
declare @cols2 nvarchar(max)
 select  @cols1 = stuff(( select distinct top 100 percent ',isnull([' + convert(varchar,t2.Calldate,105) + '],''-'')'
                           from    [Source table] as t2
                          order by 1
                            for xml path('')), 1, 1, '') + ''

 select  @cols2 = stuff(( select distinct top 100 percent ',[' + convert(varchar,t2.Calldate,105) + ']'
                           from    [Source table] as t2
                          order by 1
                            for xml path('')), 1, 1, '') + ''
declare @query nvarchar(max)
set @query = 'select [Mobile no],' + @cols1 + ' 
from (select [Mobile no],convert(varchar,CallDate,105) CallDate, Case When Disposition is NULL then ''-'' Else Disposition End as Disposition  from [Source table]) t1
               pivot (max(Disposition) for Calldate in (' + @cols2 + ')) as p'
print @query
exec (@query)
go
go
select * from [Source table]
go
drop table [source table]

Select allOpen in new window

Exactly, a3...note, you will want to repeat the column value piece or you will end up with an unnamed column in the final selection since the is null test being an expression needs to be aliased.

That worked perfectly fine. Thanks @Angel

BTW The column names have no name: the column name is supposed to be dates.
Please refer to my resultset file attached.
resultSet.xls

Thanks again :)

Glad we were able to help you!
Best regards and happy coding,
Kevin