Count distinct SPL help

[thuhuongle]

Hi, 
Looking forward to learn from you guys. I am stucked at this calculation: Total of product in contract.
I made a simple dataset to simplify my data. 

| makeresults 
| eval date = "2017-01-30" , source = "a", id="111"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source
| append 
    [| makeresults 
| eval date = "2017-01-30" , source = "b", id="222"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source]
| append 
    [| makeresults 
| eval date = "2019-08-20" , source = "a", id="333"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source]
| append 
    [| makeresults 
| eval date = "2020-01-20" , source = "a", id="444"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source]
| append 
    [| makeresults 
| eval date = "2020-03-20" , source = "b", id="555"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source]

INPUT: let's image _time is date time of buying record

_time	signed contract date	id	source
2020-10-30 14:55:55	2017-01-30	111	a
2020-10-30 14:55:55	2017-01-30	222	b
2020-08-30 14:55:55	2019-08-20	333	a
2020-01-30 14:55:55	2020-01-20	444	a
2020-09-30 14:55:55	2020-03-20	555	b

The expected output: Count total product in contract from 12/2019 to 03/2020

Time	total_nb_product	Source
12/2019	2	a
12/2019	1	b
01/2020	3	a
01/2020	1	b
02/2020	3	a
02/2020	1	b
03/2020	3	a
03/2020	2	b

Thank for your time and hope to received your suggestion. 
Great weekend

[ITWhisperer · Accepted solution]

| makeresults 
| eval date = "2017-01-30" , source = "a", id="111"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source
| append 
    [| makeresults 
| eval date = "2017-01-30" , source = "b", id="222"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source]
| append 
    [| makeresults 
| eval date = "2019-08-20" , source = "a", id="333"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source]
| append 
    [| makeresults 
| eval date = "2020-01-20" , source = "a", id="444"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source]
| append 
    [| makeresults 
| eval date = "2020-03-20" , source = "b", id="555"
| makemv delim="," id
| makemv delim="," date
| makemv delim="," source]
| eval Time=round(relative_time(strptime(date,"%Y-%m-%d"),"@mon"))
| makecontinuous Time span=1mon
| fields - _time date
| eventstats values(source) as sources
| mvexpand sources
| eval source=coalesce(source, sources)
| eval id=if(source=sources,id,null)
| fields - source
| rename sources as Source
| stats values(id) as id by Time Source
| streamstats count(id) as total_nb_product by Source
| fields Time total_nb_product Source
| fieldformat Time=strftime(Time,"%m/%Y")

[thuhuongle]

Great ideal ITWhisperer,
It is exactly what i am looking for. I am developing a sum count between date, but your solution is straightforward and much better. Hope to see you arround very soon