Hello,

I have executed the MPI implementation with the xac file as the input matrix. It finished with success but no tree was printed nor written in any file. Debugging the code by printing messages I have realized that the method printBestSolutions in uDataStr.c was only called from the method branchAndBound in uBnB.c that is only executed when num_taxa < 8 (main.c). Otherwise the method paBnB from opBnB.c is executed and no print or recover of the resulting tree can be achieved.

Even when having a matrix with num_taxa < 8 that method doesn't enter the while (bestCostStackHead != NULL) in the printBestSolutions method and so it doesn't print anything. For this purpose I used the taxons T17, T21, T3, T8, T14, T20, T0 of the xac file.

Hope you can find and implement a solution. Best regards

Hello,

I was looking the code for knowing how to use it and what implementation was used in MPI_implementation. First of all, I was surprised about watching a variable named sharedQ in the folder SMP_ListBased in files such as frameWork.c and oBnB.c that are not present in the SMP_QueueBasedSynchronous neither in the MPI_Implementation. From my comparisson of the code and accordingly to what was written in the paper I suppose that MPI_Implementation is the Queue one, but that variable was disturbing me.

Another question is about the input variables in the MPI_Implementation. The arrangements_per_level input paramenter is related to the number of arrangements of TBR for each level. The other tuning parameter mentioned in the paper (cardinality k of the set of subtrees that a processor receives from the shared queue) couldn't be inserted right? (In case it is the Queue method)

Finally, I have doubts regarding the input matrix (alignment). In the example file xac there are no N characters for undeciphered nucleotides nor gaps. This algorithm is able to handle both or some of them?

Thank you very much for sharing the code and the research!

In the main.c file you have accidentally leaved a strcat(inFile,"/tmp/"); on line 28 so it is added to the beginning of the path of the input matrix.