Sarah Lutteropp

I did some digging and searching in the slack... Some maybe relevant old Slack messages from @stamatak and @celinescornavacca about computing the number of parameters: @stamatak said: `This is more...

@celinescornavacca Here are some example values from the CSV (https://drive.google.com/file/d/1His_wi_i0Dz3E9ye6HFz6b1dhdMhnlZ8/view?usp=sharing), such that you don't have to open the table yourself: - for dataset datasets_small_network/0_5_taxa_1_reticulations_SamplingType.PERFECT_UNIFORM_SAMPLING_sampling_LikelihoodType.AVERAGE_likelihood_1000_msasize and running NetRAX starting from best...

I moved some observations about LikelihoodType.BEST to a new GitHub issue, to be please discussed there and not in this issue here: https://github.com/lutteropp/NetRAX/issues/20

At least it is pretty obvious now that the problem does not have to do with the BIC. Instead, it gets more and more clear that something is fishy with...

@celinescornavacca I cannot see any newly added columns in the CSV file. What is a reticulation insertion? I thought an arc insertion introduces a new reticulation (which means, 1 new...

These are the only number-of-reticulations-changing moves I implemented in NetRAX 

Also, I still want to double-check if the MSA is generated correctly (using the methodology I proposed here: https://github.com/lutteropp/NetRAX/issues/19#issuecomment-736420010). Because, even if I choose SamplingType.PERFECT_UNIFORM_SAMPLING (which essentially overwrites all reticulation...

It could very well be that the MSA currently generated with calling SeqGen is saying "ditch this simulated network with its multiple displayed trees, I am deciding to support only...

> This can be due to the inheritance probabilities not in the SamplingType.PERFECT_UNIFORM_SAMPLING case, where the reticulation probabilities are ignored and each reticulation is made equally likely before simulating the...

I have created a GitHub issue for the MSA discussion: https://github.com/lutteropp/NetRAX/issues/21 And here is a GitHub issue for the BIC dicussion: https://github.com/lutteropp/NetRAX/issues/22 But one note to the BIC discussion from...