Multi level cross tab dynamic query - sql

Question

I am trying to find a query that can generate cross-tab output at multiple levels dynamically. I did find few solutions online that returns dynamic cross-tab results but it returns only at single level. Below is the SQL fiddle:
CREATE TABLE dbo.PopulationDetails
(
Country VARCHAR(50),
State VARCHAR(50),
Population BIGINT,
SeatsInHouse INT
)
INSERT INTO PopulationDetails
VALUES('United States','California', 38332521, 53),
('United States','Texas', 26448193, 36),
('United States','New York', 19651127, 27),
('United States','Florida', 19552860, 27),
('United States','Illinois', 12882135, 18)
I want my output should look like below. The number of states are not fixed and these may vary as per the requirement.
United States
California Texas New York Florida Illinois
Population 38332521 26448193 19651127 19552860 12882135
SeatsInHouse 53 36 27 27 18

Answer 1

As I said in my comment, multi-level column headers can't be done via SQL. You'd have to format the data in a presentation layer like your application or SSRS. If you want to get the country and state values "together", then you'd have to concatenate the names together and make that your new column names.
If you want to get the result in SQL, I'd start by concatenating the country and state, and unpivot the columns population and SeatsInHouse first. The basic syntax for this process would be:
select
country_state = replace(pd.Country +'_'+pd.State, ' ', ''),
c.col,
c.value
from dbo.PopulationDetails pd
cross apply
(
values
('Population', pd.population),
('SeatsInHouse', pd.SeatsInHouse)
) c (col, value);
See SQL Fiddle with Demo. This gives a result:
| COUNTRY_STATE | COL | VALUE |
|-------------------------|--------------|----------|
| UnitedStates_California | Population | 38332521 |
| UnitedStates_California | SeatsInHouse | 53 |
| UnitedStates_Texas | Population | 26448193 |
| UnitedStates_Texas | SeatsInHouse | 36 |
| UnitedStates_NewYork | Population | 19651127 |
| UnitedStates_NewYork | SeatsInHouse | 27 |
You'll see that you now have two rows for each Country_State combination. You can now pivot those Country_State values into columns:
select col, UnitedStates_California, UnitedStates_Texas,
UnitedStates_NewYork, UnitedStates_Florida,
UnitedStates_Illinois
from
(
select
country_state = replace(pd.Country +'_'+pd.State, ' ', ''),
c.col,
c.value
from dbo.PopulationDetails pd
cross apply
(
values
('Population', pd.population),
('SeatsInHouse', pd.SeatsInHouse)
) c (col, value)
) d
pivot
(
max(value)
for country_state in (UnitedStates_California, UnitedStates_Texas,
UnitedStates_NewYork, UnitedStates_Florida,
UnitedStates_Illinois)
) piv;
See SQL Fiddle with Demo.
Now, if you need this done dynamically then you'd have to use dynamic SQL which creates a string that is then executed.
DECLARE #cols AS NVARCHAR(MAX),
#query AS NVARCHAR(MAX)
select #cols = STUFF((SELECT ',' + QUOTENAME(country_state)
from
(
select country_state = replace(Country +'_'+State, ' ', '')
from dbo.PopulationDetails
) d
group by country_state
order by country_state
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
set #query = 'SELECT col, ' + #cols + '
from
(
select
country_state = replace(pd.Country +''_''+pd.State, '' '', ''''),
c.col,
c.value
from dbo.PopulationDetails pd
cross apply
(
values
(''Population'', pd.population),
(''SeatsInHouse'', pd.SeatsInHouse)
) c (col, value)
) x
pivot
(
max(value)
for country_state in (' + #cols + ')
) p '
exec sp_executesql #query;
See SQL Fiddle with Demo. Both gives a result:
| COL | UNITEDSTATES_CALIFORNIA | UNITEDSTATES_FLORIDA | UNITEDSTATES_ILLINOIS | UNITEDSTATES_NEWYORK | UNITEDSTATES_TEXAS |
|--------------|-------------------------|----------------------|-----------------------|----------------------|--------------------|
| Population | 38332521 | 19552860 | 12882135 | 19651127 | 26448193 |
| SeatsInHouse | 53 | 27 | 18 | 27 | 36 |