discrete choice model in r with mlogit with nested data

#1
My data has multiple observations within participants.
Each participant has 10 trials.
For each trial, there are 4 options, there are variables associated with each of the options. The participant select one option at a time for 3 times. Each time one option is selected, it is removed and the participant need to select from the remaining ones. The variables associated with remaining choices will change. So my data looks like below. There is no subject level variables.

Since my data is nested, with repeated observations within individuals, the examples of the simple mlogit models (each person makes only one choice) online may not apply to my data. What is the appropriate way to analyze my data?

Code:
ID	Trial	C_num	option	var1		var2		choice
S1	1	1	C	0.969311934	-0.328713815	1
S1	1	1	D	0.368459569	-0.100616452	0
S1	1	1	A	0.402664843	-0.194083578	0
S1	1	1	B	0.389439842	-0.257392906	0
S1	1	2	D	0.150329117	-0.097984944	0
S1	1	2	A	0.302840263	0.209900197	1
S1	1	2	B	0.338795387	0.122258074	0
S1	1	3	D	0.216206285	-0.11890087	0
S1	1	3	B	0.297109623	0.250219284	1
S1	2	1	C	0.79171215	-0.178453098	0
S1	2	1	D	0.168236191	-0.099548578	0
S1	2	1	A	0.37572025	-0.027878373	1
S1	2	1	B	0.278136821	0.265868384	0
S1	2	2	D	0.17707435	-0.125174593	0
S1	2	2	C	0.7098889	-0.265653	0
S1	2	2	B	0.349678638	0.209010144	1
S1	2	3	D	0.174723298	-0.099354708	0
S1	2	3	C	0.808463668	-0.136589522	1
 
Last edited:
#2
So what i have tried is to treat each instance of choice as a separate unit. I know this shouldn't be appropriate since the 2nd and 3rd choice is dependent on the prior ones, you only get to choose what remains.

But I can't do a simple rank order of the 4 choices either because every time one option is selected, the variables associated with the remaining options would change to a new number, and participants are only judging using the most current state, not the ones before they made their previous decision.

I have read quite a bit on discrete choice model tutorials. They talked about variables associated with choices or decision makers, and nest modes where IIA doesn't apply. But non of them seem to apply to my data. Anyone could give me some directions, or point me to some readings that could help me find the right way for analyzing my data?