Correct usage of finding_exclusive_sets.py
Closed this issue · 1 comments
DhanjalJK commented
Kindly help me with the following questions-
- Is 'python ../find_exclusive_sets.py -mf data.json -o exc -f 5 -c 20 -v 4 -s Enumerate -ks 2 -m Saddlepoint -wf weights.npy' not the correct syntax?
- If I want to try MCMC, how to decide the value for the required parameters (--num_iterations, --num_chains, --step_length, --alpha and --mcmc_seed)?
- I used the value of --seed as 12345 for computing mutation probabilities after looking at the examples provided. What is the significance of this number and how to decide on its value?
Thanks!
matthewreyna commented
- The
find_exclusive_sets.py
script allows different choices of exclusivity tests. Please trypython ../find_exclusive_sets.py -mf data.json -o exc -f 5 -c 20 -v 4 -ks 2 -s Enumerate WRE -m Saddlepoint -wf weights.npy
to run the WR-exclusivity test from the WExT paper. Also, if you encounter further errors, please share the error message to make it easier to provide feedback. - You can test these values experimentally to see if the MCMC appears to converge. For more information about the MCMC, please see the CoMEt paper.
- The
-s
(--seed
) argument for thecompute_mutation_probabilities.py
script is a seed for the random number generator used in this script. By providing a seed, you can generate the same results when you run this script on the same data at different times. You can choose any integer.