Hi Everyone,

I'm trying to do some predictive forecasting with parameters but I'm coming across an issue when using parameters into the future. What i'm trying to do is forecast the amount of people on a waiting list in the future but this can be altered by using parameters which will remove more than forecast. So for example, I have 100 people on a list in Jan 150 in Feb 200 in March etc, and I can forecast these out fine, I can even add in the parameter to for a what if (Extra Clinic), but the problem with the what if is that if I change the what if variable, it changes all the past information too, and I only want to change the future forecast.

The data contains a record for each person on the list so for January I have 100 records and so on, so I created a distinct count measure of patients (No. Of Patients) to get me the total number in that month.
I found this formula which kind of works but instead of taking the distinct count of the values at the point it takes all the records. The sn_cal_month_id is the date rage.
Here's the query:


WL Prediction Parameter = SUMX(Query1, IF((Query1[sn_cal_month_id]) > 201801, 'Query1'[No. Of Patients] - 'Extra Clinic'[Extra Clinic Value], Query1[No. Of Patients]))

So instead of getting just a figure removed off a single month, it removes the entire amount of records from the graph.
Any idea what I'm doing so horribly wrong?
Thanks,