How to handle multi-treatment use-case with DynamicDML?

[meprem]

Hi Team,

We have 50+ control variables and N (>5) different types of treatments applied in each time-period. Our goal is to rank these treatments (O1…O7) based on their effectiveness using DynamicDML : https://github.com/py-why/EconML/blob/main/notebooks/Dynamic%20Double%20Machine%20Learning%20Examples.ipynb.

Can DynamicDML class handle multiple different treatments in each time-period? If yes, could you please guide me to relevant code? Thank you.

Here is the sample data:

<style> </style>

Company	Investment Type	Year	Features	Investment	Revenue
A	O1	2018	...	$1,000	$10,000
A	O2	2018	...	$2,000	$12,000
A	O3	2018	...	$800	$15,000
A	O1	2019	...	$1,000	$10,000
A	O2	2019	...	$2,000	$12,000
A	O4	2019	...	$3,000	$15,000
A	O4	2020	...	$1,000	$10,000
A	O5	2020	...	$2,000	$12,000
A	O6	2020	...	$3,000	$15,000
B	O1	2018	...	$1,000	$10,000
B	O2	2018	...	$2,000	$12,000
B	O3	2018	...	$800	$15,000
B	O1	2019	...	$1,000	$10,000
B	O2	2019	...	$2,000	$12,000
B	O4	2019	...	$3,000	$15,000
B	O4	2020	...	$1,000	$10,000
B	O5	2020	...	$2,000	$12,000
B	O6	2020	...	$3,000	$15,000
B	O7	2020	...	$3,000	$15,000

https://github.com/py-why/EconML/blob/main/notebooks/Dynamic%20Double%20Machine%20Learning%20Examples.ipynb

Regards,
Prem