sql find max value of a column and select as new field - sql

Question

I am currently working with some ambulance data and in some scenario's multiple vehicles are dispatched to the same call. Of those dispatched sometimes they are stood down(diverted to a different call.)
The relevent fields are incidentID, VehicleAllocationSequenceNumber snd VehicleArrivalAtSceneDateTime.
An incidentID can map to multiple rows with different VehicleAllocationSequenceNumber and VehicleArrivalAtSceneDateTime.
VehicleAllocationSequenceNumber defines the order in which the vehicles arrived on the scene. VehicleArrivalAtSceneDateTime is either a datetime or NULL if the vehicle never arrived.
I want to select distinct IncidentID's with and condense the other columns into two new features; Number_of_Vehicles_assigned and number_of_vehicles_on_scene
I am fairly new to queries of this complexity and I am looking for a way to do this without using a cursor.
Thanks in advance
Edit:
Nevermind I nailed the query, think I'm just a bit scatter brained this morning. Here is my query for posterity's sake.
Feedback is welcome:
select
IncidentID,
max(VehicleAllocationSequenceNumber) as number_vehicles_assigned,
count(VehicleArrivalAtSceneDateTime) as number_of_vehicles_on_scene
FROM
[dwuserobj].[acc].[AmbulanceLinkedDatasetForModelling]
where
IncidentID in(
select distinct
IncidentID
from
[dwuserobj].[acc].[AmbulanceLinkedDatasetForModelling]
)
and vehicleArrivalAtSceneDateTime is not null
group by
IncidentID;

Answer 1

Use a case with a not exists
select distinct incidentID,
case
when not exists
(
select 1
from AmbulanceTable t2
where t2.incidentID = t1.incidentID
and t2.VehicleAllocationSequenceNumber is null
) then 1
else 0 as alloc,
case
when not exists
(
select 1
from AmbulanceTable t3
where t3.incidentID = t1.incidentID
and t3.VehicleArrivalAtSceneDateTime is null
) then 1
else 0 as onscene
from AmbulanceTable t1
Edit:
So no_vehicles is number of vehicles... my bad:
select incidentid
count(VehicleAllocationSequenceNumber) as num_assigned,
count(VehicleArrivalAtSceneDateTime ) as num_oncscene
from AmbulanceLinkedDatasetForModelling
group by incidentid
this will give you zeros if none assigned/on scene

Answer 2

this should provide you with the logic you need. I made a temp table so you can see how the incidentID interacts with the AllocationSequenceNumber and the SceneDateTime.
DECLARE #AmbulanceLinkedDatasetForModelling TABLE (IncidentID int, VehicleAllocationSequenceNumber int, VehicleArrivalAtSceneDateTime datetime)
INSERT INTO #AmbulanceLinkedDatasetForModelling
SELECT 1, 500, NULL UNION ALL -- the vehicle never arrived
SELECT 2, 501, '2016-05-20 06:07:00.370' UNION ALL
SELECT 3, 502, '2018-01-05 08:15:08.970' UNION ALL
SELECT 4, 503, NULL UNION ALL -- the vehicle never arrived
SELECT 4, 504, '2017-04-11 11:22:01.360'UNION ALL
SELECT 4, 505, '2017-04-11 11:32:01.360'
SELECT
IncidentID,
number_vehicles_assigned=COUNT(DISTINCT VehicleAllocationSequenceNumber),
no_vehicles_on_scene=SUM(CASE WHEN VehicleArrivalAtSceneDateTime IS NOT NULL THEN 1 ELSE 0 END)
FROM
#AmbulanceLinkedDatasetForModelling
GROUP BY
IncidentID