search

NBISweden / MrBayes

MrBayes is a program for Bayesian inference and model choice across a wide range of phylogenetic and evolutionary models. For documentation and downloading the program, please see the home page:
http://NBISweden.github.io/MrBayes/
GNU General Public License v3.0
233 stars 80 forks source link

mpirun does not run the parallel version #50

Closed Pejvak1 closed 5 years ago

Pejvak1 commented 7 years ago

Hi All,

I am running mpi version of mrbayes on a cluster, but when I use the standard command, unlike on my own laptop, it gives me some weird output. I don't really know what is going on, but, it seems that instead of running one of every 32 chains (16 chains for each run) on one core, it is running 16 chains (or 32, I'm not sure) on each core.

  1. When the mrbayes start-up message comes up it "does not acknowledge a parallel version", like it does on my laptop. The following image is what I get when I run it on cluster:

image)

This is what I get on my laptop:

image

  1. The program runs either once with the taxa list being unusually very very long, with each taxon repeated many times, or it runs many times (and subsequently I get the uploaded image above and the follow-up matrix reading many many times), before it gets to the standard active parameter section. Either case is dependent on how I run the program though (please see below). Also the order of taxa is weird. It looks like the following:

image

and the following:

image

  1. Once it does get to the active parameters similarly everything that I would see once on the screen, I see multiple times, including the log-likelihood for each chain in each run. See the image:

image

  1. Finally, each generation is printed multiple times (probably equivalent to the number of cores), i.e. I get: 5000 [xxxxx] (xxxx) (xxxx) (xxxxx) (xxxx) (xxxxx).. time multiple times. The running time for appx. 1000 seqs, each 1000 residues long, on 10mil gens is predicted by mb to be about 13000 hours on most chains, and 30000 on chain!

  1. mpirun and everything necessary for running parallel mb, on the cluster, was installed by the systems administrator. On my laptop I obviously installed everything with sudo, so I didn't have to set any paths, which brings me to the second difference.

  2. On laptop, All I do is mpirun -np 8 mb <file> , whereas on the cluster the following is what goes in my script:

set path = ( $path /usr/lib64/openmpi/bin ) setenv LD_LIBRARY_PATH /usr/local/lib /usr/lib64/openmpi/bin/mpirun -np 2 /biol/programs/bin/mb primates.nex

After getting the weird results, I did however change the way I ran mb. So instead of using: ../bin/mb

I used: ../bin/mb_3.2.6_64bit_mpi

But instead of mb running many times, it ran once, but everything was repeated many times.

I would really appreciate any help about this.

Cheers, Pej.

nylander commented 6 years ago

I can't repeat this behavior on the current code using mpirun or submitting parallel jobs through a queue system (slurm).

ferroao commented 6 years ago

In my case it runs many times. 

                        MrBayes v3.2.2 x64

                  (Bayesian Analysis of Phylogeny)

          Distributed under the GNU General Public License

           Type "help" or "help <command>" for information
                 on the commands that are available.

               Type "about" for authorship and general
                   information about the program.

Executing file "primates.nex" UNIX line termination Longest line length = 924 Parsing file Expecting NEXUS formatted file Reading data block Allocated taxon set Allocated matrix Defining new matrix with 12 taxa and 898 characters Data is Dna Data matrix is not interleaved Gaps coded as - Taxon 1 -> Tarsius_syrichta Taxon 2 -> Lemur_catta Taxon 3 -> Homo_sapiens Taxon 4 -> Pan Taxon 5 -> Gorilla Taxon 6 -> Pongo Taxon 7 -> Hylobates Taxon 8 -> Macaca_fuscata Taxon 9 -> M_mulatta Taxon 10 -> M_fascicularis Taxon 11 -> M_sylvanus Taxon 12 -> Saimiri_sciureus Successfully read matrix Setting default partition (does not divide up characters) Setting model defaults Seed (for generating default start values) = 1533850978 Setting output file names to "primates.nex.run.<p|t>" Setting Nst to 6 Setting Rates to Invgamma Successfully set likelihood model parameters

                        MrBayes v3.2.2 x64

                  (Bayesian Analysis of Phylogeny)

          Distributed under the GNU General Public License

           Type "help" or "help <command>" for information
                 on the commands that are available.

               Type "about" for authorship and general
                   information about the program.

Executing file "primates.nex" UNIX line termination Longest line length = 924 Parsing file Expecting NEXUS formatted file Reading data block Allocated taxon set Allocated matrix Defining new matrix with 12 taxa and 898 characters Data is Dna Data matrix is not interleaved Gaps coded as - Taxon 1 -> Tarsius_syrichta Taxon 2 -> Lemur_catta Taxon 3 -> Homo_sapiens Taxon 4 -> Pan Taxon 5 -> Gorilla Taxon 6 -> Pongo Taxon 7 -> Hylobates Taxon 8 -> Macaca_fuscata Taxon 9 -> M_mulatta Taxon 10 -> M_fascicularis Taxon 11 -> M_sylvanus Taxon 12 -> Saimiri_sciureus Successfully read matrix Setting default partition (does not divide up characters) Setting model defaults Seed (for generating default start values) = 1533850978 Setting output file names to "primates.nex.run.<p|t>" Setting Nst to 6 Setting Rates to Invgamma Successfully set likelihood model parameters

                        MrBayes v3.2.2 x64

                  (Bayesian Analysis of Phylogeny)

          Distributed under the GNU General Public License

           Type "help" or "help <command>" for information
                 on the commands that are available.

               Type "about" for authorship and general
                   information about the program.

Executing file "primates.nex" UNIX line termination Longest line length = 924 Parsing file Expecting NEXUS formatted file Reading data block Allocated taxon set Allocated matrix Defining new matrix with 12 taxa and 898 characters Data is Dna Data matrix is not interleaved Gaps coded as - Taxon 1 -> Tarsius_syrichta Taxon 2 -> Lemur_catta Taxon 3 -> Homo_sapiens Taxon 4 -> Pan Taxon 5 -> Gorilla Taxon 6 -> Pongo Taxon 7 -> Hylobates Taxon 8 -> Macaca_fuscata Taxon 9 -> M_mulatta Taxon 10 -> M_fascicularis Taxon 11 -> M_sylvanus Taxon 12 -> Saimiri_sciureus Successfully read matrix Setting default partition (does not divide up characters) Setting model defaults Seed (for generating default start values) = 1533850978

                        MrBayes v3.2.2 x64

                  (Bayesian Analysis of Phylogeny)

          Distributed under the GNU General Public License

           Type "help" or "help <command>" for information
                 on the commands that are available.

               Type "about" for authorship and general
                   information about the program.

Executing file "primates.nex" UNIX line termination Longest line length = 924 Parsing file Expecting NEXUS formatted file Reading data block Allocated taxon set Allocated matrix Defining new matrix with 12 taxa and 898 characters Data is Dna Data matrix is not interleaved Gaps coded as - Taxon 1 -> Tarsius_syrichta Taxon 2 -> Lemur_catta Taxon 3 -> Homo_sapiens Taxon 4 -> Pan Taxon 5 -> Gorilla Taxon 6 -> Pongo Taxon 7 -> Hylobates Taxon 8 -> Macaca_fuscata Taxon 9 -> M_mulatta Setting output file names to "primates.nex.run.<p|t>" Setting Nst to 6 Setting Rates to Invgamma Successfully set likelihood model parameters Taxon 10 -> M_fascicularis Taxon 11 -> M_sylvanus Taxon 12 -> Saimiri_sciureus Setting number of generations to 1000 Setting sample frequency to 10 Running Markov chain MCMC stamp = 4944105845 Seed = 1073401300 Successfully read matrix Swapseed = 1533850978 Setting default partition (does not divide up characters) Setting model defaults Seed (for generating default start values) = 1533850978 Model settings:

     Data not partitioned --
        Datatype  = DNA
        Nucmodel  = 4by4
        Nst       = 6
                    Substitution rates, expressed as proportions
                    of the rate sum, have a Dirichlet prior
                    (1.00,1.00,1.00,1.00,1.00,1.00)
        Covarion  = No
        # States  = 4
                    State frequencies have a Dirichlet prior
                    (1.00,1.00,1.00,1.00)
        Rates     = Invgamma
                    Gamma shape parameter is exponentially
                    distributed with parameter (2.00).
                    Proportion of invariable sites is uniformly dist-
  Setting number of generations to 1000
  Setting sample frequency to 10
  Running Markov chain
  MCMC stamp = 4944105845
  Seed = 1073401300
  Swapseed = 1533850978
  Model settings:

     Data not partitioned --
        Datatype  = DNA
        Nucmodel  = 4by4
        Nst       = 6
                    Substitution rates, expressed as proportions
                    of the rate sum, have a Dirichlet prior
                    (1.00,1.00,1.00,1.00,1.00,1.00)
        Covarion  = No
        # States  = 4
                    State frequencies have a Dirichlet prior
                    (1.00,1.00,1.00,1.00)
        Rates     = Invgamma
                    Gamma shape parameter is exponentially
                    distributed with parameter (2.00).
                    Proportion of invariable sites is uniformly dist-
                    ributed on the interval (0.00,1.00).
                    Gamma distribution is approximated using 4 categories.
                    Likelihood summarized over all rate categories in each generation.

  Active parameters: 

     Parameters
     ------------------
     Revmat           1
     Statefreq        2
     Shape            3
     Pinvar           4
     Ratemultiplier   5
     Topology         6
     Brlens           7
     ------------------

     1 --  Parameter  = Revmat
           Type       = Rates of reversible rate matrix
           Prior      = Dirichlet(1.00,1.00,1.00,1.00,1.00,1.00)

     2 --  Parameter  = Pi
           Type       = Stationary state frequencies
           Prior      = Dirichlet

     3 --  Parameter  = Alpha
           Type       = Shape of scaled gamma distribution of site rates
           Prior      = Exponential(2.00)

     4 --  Parameter  = Pinvar
           Type       = Proportion of invariable sites
           Prior      = Uniform(0.00,1.00)

     5 --  Parameter  = Ratemultiplier
           Type       = Partition-specific rate multiplier
           Prior      = Fixed(1.0)

     6 --  Parameter  = Tau
           Type       = Topology
           Prior      = All topologies equally probable a priori
           Subparam.  = V

     7 --  Parameter  = V
           Type       = Branch lengths
           Prior      = Unconstrained:Exponential(10.0)

  The MCMC sampler will use the following moves:
     With prob.  Chain will use move
        1.09 %   Dirichlet(Revmat)
        1.09 %   Slider(Revmat)
        1.09 %   Dirichlet(Pi)
        1.09 %   Slider(Pi)
        2.17 %   Multiplier(Alpha)
        2.17 %   Slider(Pinvar)
       10.87 %   ExtSPR(Tau,V)
       10.87 %   ExtTBR(Tau,V)
       10.87 %   NNI(Tau,V)
       10.87 %   ParsSPR(Tau,V)
       32.61 %   Multiplier(V)
       10.87 %   Nodeslider(V)
        4.35 %   TLMultiplier(V)

                    ributed on the interval (0.00,1.00).
                    Gamma distribution is approximated using 4 categories.
                    Likelihood summarized over all rate categories in each generation.

  Setting number of generations to 1000
  Setting sample frequency to 10
  Running Markov chain
  MCMC stamp = 4944105845
  Seed = 1073401300
  Swapseed = 1533850978
  Model settings:

     Data not partitioned --
        Datatype  = DNA
        Nucmodel  = 4by4
        Nst       = 6
                    Substitution rates, expressed as proportions
                    of the rate sum, have a Dirichlet prior
                    (1.00,1.00,1.00,1.00,1.00,1.00)
        Covarion  = No
        # States  = 4
                    State frequencies have a Dirichlet prior
                    (1.00,1.00,1.00,1.00)
        Rates     = Invgamma
                    Gamma shape parameter is exponentially
                    distributed with parameter (2.00).
                    Proportion of invariable sites is uniformly dist-
                    ributed on the interval (0.00,1.00).
                    Gamma distribution is approximated using 4 categories.
                    Likelihood summarized over all rate categories in each generation.

  Active parameters: 

     Parameters
     ------------------
     Revmat           1
     Statefreq        2
     Shape            3
     Pinvar           4
     Ratemultiplier   5
     Topology         6
     Brlens           7
     ------------------
  Division 1 has 413 unique site patterns
  Initializing conditional likelihoods

     1 --  Parameter  = Revmat
           Type       = Rates of reversible rate matrix
           Prior      = Dirichlet(1.00,1.00,1.00,1.00,1.00,1.00)

     2 --  Parameter  = Pi
           Type       = Stationary state frequencies
  Using standard SSE likelihood calculator for division 1 (single-precision)
           Prior      = Dirichlet

     3 --  Parameter  = Alpha
           Type       = Shape of scaled gamma distribution of site rates
           Prior      = Exponential(2.00)

     4 --  Parameter  = Pinvar
           Type       = Proportion of invariable sites
           Prior      = Uniform(0.00,1.00)

     5 --  Parameter  = Ratemultiplier
           Type       = Partition-specific rate multiplier
           Prior      = Fixed(1.0)

     6 --  Parameter  = Tau
           Type       = Topology
           Prior      = All topologies equally probable a priori
           Subparam.  = V

     7 --  Parameter  = V
           Type       = Branch lengths
           Prior      = Unconstrained:Exponential(10.0)

  The MCMC sampler will use the following moves:
     With prob.  Chain will use move
        1.09 %   Dirichlet(Revmat)
        1.09 %   Slider(Revmat)
        1.09 %   Dirichlet(Pi)
        1.09 %   Slider(Pi)
        2.17 %   Multiplier(Alpha)
        2.17 %   Slider(Pinvar)
       10.87 %   ExtSPR(Tau,V)
       10.87 %   ExtTBR(Tau,V)
       10.87 %   NNI(Tau,V)
       10.87 %   ParsSPR(Tau,V)
       32.61 %   Multiplier(V)
       10.87 %   Nodeslider(V)
        4.35 %   TLMultiplier(V)

  Setting output file names to "primates.nex.run<i>.<p|t>"
  Setting Nst to 6
  Setting Rates to Invgamma
  Successfully set likelihood model parameters
  Active parameters: 

     Parameters
     ------------------
     Revmat           1
     Statefreq        2
     Shape            3
     Pinvar               Division 1 has 413 unique site patterns
  Initializing conditional likelihoods
  Using standard SSE likelihood calculator for division 1 (single-precision)

4 Ratemultiplier 5 Topology 6 Brlens 7

     1 --  Parameter  = Revmat
           Type       = Rates of reversible rate matrix
           Prior      = Dirichlet(1.00,1.00,1.00,1.00,1.00,1.00)

     2 --  Parameter  = Pi
           Type       = Stationary state frequencies
           Prior      = Dirichlet

     3 --  Parameter  = Alpha
           Type       = Shape of scaled gamma distribution of site rates
           Prior      = Exponential(2.00)

     4 --  Parameter  = Pinvar
           Type       = Proportion of invariable sites
           Prior      = Uniform(0.00,1.00)

     5 --  Parameter  = Ratemultiplier
           Type       = Partition-specific rate multiplier
           Prior      = Fixed(1.0)

     6 --  Parameter  = Tau
           Type       = Topology
           Prior      = All topologies equally probable a priori
           Subparam.  = V

     7 --  Parameter  = V
           Type       = Branch lengths
           Prior      = Unconstrained:Exponential(10.0)

  The MCMC sampler will use the following moves:
     With prob.  Chain will use move
        1.09 %   Dirichlet(Revmat)
        1.09 %   Slider(Revmat)
        1.09 %   Dirichlet(Pi)
        1.09 %   Slider(Pi)
        2.17 %   Multiplier(Alpha)
  Setting number of generations to 1000
  Setting sample frequency to 10
  Running Markov chain
  MCMC stamp = 4944105845
  Seed = 1073401300
  Swapseed = 1533850978
  Model settings:

     Data not partitioned --
        Datatype  = DNA
        Nucmodel  = 4by4
        Nst       = 6
                    Substitution rates, expressed as proportions
                    of the rate sum, have a Dirichlet prior
        2.17 %   Slider(Pinvar)
                    (1.00,1.00,1.00,1.00,1.00,1.00)
        Covarion  = No
        # States  = 4
                    State frequencies have a Dirichlet prior
                    (1.00,1.00,1.00,1.00)
        Rates     = Invgamma
                    Gamma shape parameter is exponentially
                    distributed with parameter (2.00).
                    Proportion of invariable sites is uniformly dist-
                    ributed on the interval (0.00,1.00).
                    Gamma distribution is approximated using 4 categories.
                    Likelihood summarized over all rate categories in each generation.

  Active parameters: 

     Parameters
     ------------------
     Revmat           1
     Statefreq        2
     Shape            3
     Pinvar           4
     Ratemultiplier   5
     Topology         6
     Brlens           7
     ------------------

     1 --  Parameter  = Revmat
           Type       = Rates of reversible rate matrix
           Prior      = Dirichlet(1.00,1.00,1.00,1.00,1.00,1.00)

     2 --  Parameter  = Pi
           Type       = Stationary state frequencies
           Prior      = Dirichlet

     3 --  Parameter  = Alpha
           Type       = Shape of scaled gamma distribution of site rates
           Prior      = Exponential(2.00)

     4 --  Parameter  = Pinvar
           Type       = Proportion of invariable sites
           Prior      = Uniform(0.00,1.00)

     5 --  Parameter  = Ratemultiplier
           Type       = Partition-specific rate multiplier
           Prior      = Fixed(1.0)

     6 --  Parameter  = Tau
           Type       = Topology
           Prior      = All topologies equally probable a priori
           Subparam.  = V

     7 --  Parameter  = V
           Type       = Branch lengths
           Prior      = Unconstrained:Exponential(10.0)

  The MCMC sampler will use the following moves:
     With prob.  Chain will use move
        1.09 %   Dirichlet(Revmat)
        1.09 %   Slider(Revmat)
        1.09 %   Dirichlet(Pi)
        1.09 %   Slider(Pi)
        2.17 %   Multiplier(Alpha)
        2.17 %   Slider(Pinvar)
       10.87 %   ExtSPR(Tau,V)
       10.87 %   ExtTBR(Tau,V)
       10.87 %   NNI(Tau,V)
       10.87 %   ParsSPR(Tau,V)
       32.61 %   Multiplier(V)
       10.87 %   Nodeslider(V)
        4.35 %   TLMultiplier(V)

       10.87 %   ExtSPR(Tau,V)
       10.87 %   ExtTBR(Tau,V)
       10.87 %   NNI(Tau,V)
       10.87 %   ParsSPR(Tau,V)
       32.61 %   Multiplier(V)
       10.87 %   Nodeslider(V)
        4.35 %   TLMultiplier(V)

  Division 1 has 413 unique site patterns
  Initializing conditional likelihoods
  Using standard SSE likelihood calculator for division 1 (single-precision)
  Division 1 has 413 unique site patterns
  Initializing conditional likelihoods
  Using standard SSE likelihood calculator for division 1 (single-precision)
  Initializing invariable-site conditional likelihoods

  Initial log likelihoods and log prior probs for run 1:
  Initializing invariable-site conditional likelihoods
     Chain 1 -- -7712.247222 -- -5.673047

  Initial log likelihoods and log prior probs for run 1:
     Chain 2 -- -8229.640729 -- -5.673047
     Chain 1 -- -7712.247222 -- -5.673047
     Chain 2 -- -8229.640729 -- -5.673047
     Chain 3 -- -7853.584129 -- -5.673047
     Chain 3 -- -7853.584129 -- -5.673047
     Chain 4 -- -8160.378257 -- -5.673047

  Initial log likelihoods and log prior probs for run 2:
     Chain 4 -- -8160.378257 -- -5.673047

  Initial log likelihoods and log prior probs for run 2:
     Chain 1 -- -8068.646012 -- -5.673047
  Initializing invariable-site conditional likelihoods

  Initial log likelihoods and log prior probs for run 1:
     Chain 1 -- -8068.646012 -- -5.673047
  Initializing invariable-site conditional likelihoods

  Initial log likelihoods and log prior probs for run 1:
     Chain 1 -- -7712.247222 -- -5.673047
     Chain 2 -- -8101.650142 -- -5.673047
     Chain 2 -- -8101.650142 -- -5.673047
     Chain 1 -- -7712.247222 -- -5.673047
     Chain 2 -- -8229.640729 -- -5.673047
     Chain 3 -- -8260.721019 -- -5.673047
     Chain 3 -- -8260.721019 -- -5.673047
     Chain 2 -- -8229.640729 -- -5.673047
     Chain 4 -- -7889.979133 -- -5.673047

     Chain 3 -- -7853.584129 -- -5.673047
  Overwriting file "primates.nex.run1.p"
  Overwriting file "primates.nex.run1.t"
  Overwriting file "primates.nex.run2.p"
  Overwriting file "primates.nex.run2.t"
  Overwriting file "primates.nex.mcmc"

  Using a relative burnin of 25.0 % for diagnostics

  Chain results (1000 generations requested):

      0 -- [-7712.247] (-8229.641) (-7853.584) (-8160.378) * [-8068.646] (-8101.650) (-8260.721) (-7889.979) 
     Chain 4 -- -7889.979133 -- -5.673047

     Chain 4 -- -8160.378257 -- -5.673047

  Initial log likelihoods and log prior probs for run 2:
     Chain 3 -- -7853.584129 -- -5.673047
  Overwriting file "primates.nex.run1.p"
     Chain 1 -- -8068.646012 -- -5.673047
     Chain 4 -- -8160.378257 -- -5.673047

  Initial log likelihoods and log prior probs for run 2:
     Chain 2 -- -8101.650142 -- -5.673047
     Chain 1 -- -8068.646012 -- -5.673047
     Chain 3 -- -8260.721019 -- -5.673047
     Chain 2 -- -8101.650142 -- -5.673047
  Overwriting file "primates.nex.run1.t"
     Chain 4 -- -7889.979133 -- -5.673047

  Overwriting file "primates.nex.run1.p"
  Overwriting file "primates.nex.run1.t"
     Chain 3 -- -8260.721019 -- -5.673047
     Chain 4 -- -7889.979133 -- -5.673047

  Overwriting file "primates.nex.run1.p"
  Overwriting file "primates.nex.run1.t"
  Overwriting file "primates.nex.run2.p"
  Overwriting file "primates.nex.run2.p"
  Overwriting file "primates.nex.run2.p"
  Overwriting file "primates.nex.run2.t"
  Overwriting file "primates.nex.run2.t"
  Overwriting file "primates.nex.run2.t"
  Overwriting file "primates.nex.mcmc"
  Overwriting file "primates.nex.mcmc"
  Overwriting file "primates.nex.mcmc"

  Using a relative burnin of 25.0 % for diagnostics

  Chain results (1000 generations requested):

      0 -- [-7712.247] (-8229.641) 
  Using a relative burnin of 25.0 % for diagnostics

(-7853.584) (-8160.378) * [-8068.646] (-8101.650) (-8260.721) (-7889.979) 

  Chain results (1000 generations requested):

      0 -- [-7712.247]

(-8229.641) Using a relative burnin of 25.0 % for diagnostics (-7853.584) (-8160.378) * Chain results (1000 generations requested):

[-8068.646] 0 -- (-8101.650) [-7712.247] (-8260.721) (-8229.641) (-7889.979) (-7853.584) (-8160.378) [-8068.646] (-8101.650) (-8260.721) (-7889.979) 500 -- (-6233.832) (-6145.821) [-5991.716] (-6202.195) (-6103.926) (-6213.041) [-6098.140] (-6101.498) -- 0:00:01 500 -- (-6233.832) (-6145.821) [-5991.716] (-6202.195) (-6103.926) (-6213.041) [-6098.140] (-6101.498) -- 0:00:01 500 -- (-6233.832) (-6145.821) [-5991.716] (-6202.195) (-6103.926) (-6213.041) [-6098.140] (-6101.498) -- 0:00:01 500 -- (-6233.832) (-6145.821) [-5991.716] (-6202.195) (-6103.926) (-6213.041) [-6098.140] (-6101.498) -- 0:00:01 1000 -- (-6102.739) (-5942.009) [-5903.369] (-6034.000) [-5909.130] (-6047.537) (-5973.908) (-5969.691) -- 0:00:00

  Average standard deviation of split frequencies: 0.006203
  Final log likelihoods and log prior probs for run 1 (stored and calculated):
     Chain 1 -- -6102.739461 -- 15.229932
     Chain 1 -- -6102.739480 -- 15.229932
     Chain 2 -- -5942.009386 -- 14.523801
     Chain 2 -- -5942.009363 -- 14.523801
     Chain 3 -- -5903.369331 -- 13.945189
     Chain 3 -- -5903.369429 -- 13.945189
     Chain 4 -- -6033.999928 -- 14.421609
     Chain 4 -- -6033.999872 -- 14.421609
  Final log likelihoods and log prior probs for run 2 (stored and calculated):
     Chain 1 -- -5909.129855 -- 13.061768
     Chain 1 -- -5909.129829 -- 13.061768
     Chain 2 -- -6047.537238 -- 13.588080
     Chain 2 -- -6047.537238 -- 13.588080
     Chain 3 -- -5973.908051 -- 14.274808
     Chain 3 -- -5973.907985 -- 14.274808
     Chain 4 -- -5969.690943 -- 15.615901
     Chain 4 -- -5969.690943 -- 15.615901

  Analysis completed in 1 second
  Analysis used 1.36 seconds of CPU time
  Likelihood of best state for "cold" chain of run 1 was -5900.63
  Likelihood of best state for "cold" chain of run 2 was -5941.80

  Acceptance rates for the moves in the "cold" chain of run 1:
     With prob.   (last 100)   chain accepted proposals by move
         NA           NA       Dirichlet(Revmat)
         NA           NA       Slider(Revmat)
         NA           NA       Dirichlet(Pi)
         NA           NA       Slider(Pi)
         NA           NA       Multiplier(Alpha)
         NA           NA       Slider(Pinvar)
         NA           NA       ExtSPR(Tau,V)
         NA           NA       ExtTBR(Tau,V)
         0.9 %     (  1 %)     NNI(Tau,V)
         6.3 %     (  7 %)     ParsSPR(Tau,V)
        44.2 %     ( 38 %)     Multiplier(V)
        26.0 %     ( 25 %)     Nodeslider(V)
         NA           NA       TLMultiplier(V)

  Acceptance rates for the moves in the "cold" chain of run 2:
     With prob.   (last 100)   chain accepted proposals by move
         NA           NA       Dirichlet(Revmat)
         NA           NA       Slider(Revmat)
         NA           NA       Dirichlet(Pi)
         NA           NA       Slider(Pi)
         NA           NA       Multiplier(Alpha)
         NA           NA       Slider(Pinvar)
         4.2 %     (  5 %)     ExtSPR(Tau,V)
         1.7 %     (  2 %)     ExtTBR(Tau,V)
         4.5 %     (  5 %)     NNI(Tau,V)
         8.8 %     (  9 %)     ParsSPR(Tau,V)
        48.1 %     ( 42 %)     Multiplier(V)
        23.6 %     ( 22 %)     Nodeslider(V)
         NA           NA       TLMultiplier(V)

  Chain swap information for run 1:

             1     2     3     4 
       --------------------------
     1 |        0.03  0.01  0.00 
     2 |   178        0.16  0.05 
     3 |   169   149        0.31 
     4 |   166   177   161       

  Chain swap information for run 2:

             1     2     3     4 
       --------------------------
     1 |        0.19  0.04  0.01 
     2 |   162        0.39  0.02 
     3 |   190   152        0.15 
     4 |   177   170   149       

  Upper diagonal: Proportion of successful state exchanges between chains
  Lower diagonal: Number of attempted state exchanges between chains

  Chain information:

    ID -- Heat 
   -----------
     1 -- 1.00  (cold chain)
     2 -- 0.91 
     3 -- 0.83 
     4 -- 0.77 

  Heat = 1 / (1 + T * (ID - 1))
     (where T = 0.10 is the temperature and ID is the chain number)

  Reached end of file

Tasks completed, exiting program because mode is noninteractive To return control to the command line after completion of file processing, set mode to interactive with 'mb -i ' (i is for interactive) or use 'set mode=interactive'

   1000 -- (-6102.739) (-5942.009) [-5903.369] (-6034.000) * [-5909.130] (-6047.537) (-5973.908) (-5969.691) -- 0:00:00

  Average standard deviation of split frequencies: 0.006203
  Final log likelihoods and log prior probs for run 1 (stored and calculated):
     Chain 1 -- -6102.739461 -- 15.229932
     Chain 1 -- -6102.739480 -- 15.229932
     Chain 2 -- -5942.009386 -- 14.523801
     Chain 2 -- -5942.009363 -- 14.523801
     Chain 3 -- -5903.369331 -- 13.945189
     Chain 3 -- -5903.369429 -- 13.945189
     Chain 4 -- -6033.999928 -- 14.421609
     Chain 4 -- -6033.999872 -- 14.421609
  Final log likelihoods and log prior probs for run 2 (stored and calculated):
     Chain 1 -- -5909.129855 -- 13.061768
     Chain 1 -- -5909.129829 -- 13.061768
     Chain 2 -- -6047.537238 -- 13.588080
     Chain 2 -- -6047.537238 -- 13.588080
     Chain 3 -- -5973.908051 -- 14.274808
     Chain 3 -- -5973.907985 -- 14.274808
     Chain 4 -- -5969.690943 -- 15.615901
     Chain 4 -- -5969.690943 -- 15.615901

  Analysis completed in 1 second
  Analysis used 1.37 seconds of CPU time
  Likelihood of best state for "cold" chain of run 1 was -5900.63
  Likelihood of best state for "cold" chain of run 2 was -5941.80

  Acceptance rates for the moves in the "cold" chain of run 1:
     With prob.   (last 100)   chain accepted proposals by move
         NA           NA       Dirichlet(Revmat)
         NA           NA       Slider(Revmat)
         NA           NA       Dirichlet(Pi)
         NA           NA       Slider(Pi)
         NA           NA       Multiplier(Alpha)
         NA           NA       Slider(Pinvar)
         NA           NA       ExtSPR(Tau,V)
         NA           NA       ExtTBR(Tau,V)
         0.9 %     (  1 %)     NNI(Tau,V)
         6.3 %     (  7 %)     ParsSPR(Tau,V)
        44.2 %     ( 38 %)     Multiplier(V)
        26.0 %     ( 25 %)     Nodeslider(V)
         NA           NA       TLMultiplier(V)

  Acceptance rates for the moves in the "cold" chain of run 2:
     With prob.   (last 100)   chain accepted proposals by move
         NA           NA       Dirichlet(Revmat)
         NA           NA       Slider(Revmat)
         NA           NA       Dirichlet(Pi)
         NA           NA       Slider(Pi)
         NA           NA       Multiplier(Alpha)
         NA           NA       Slider(Pinvar)
         4.2 %     (  5 %)     ExtSPR(Tau,V)
         1.7 %     (  2 %)     ExtTBR(Tau,V)
         4.5 %     (  5 %)     NNI(Tau,V)
         8.8 %     (  9 %)     ParsSPR(Tau,V)
        48.1 %     ( 42 %)     Multiplier(V)
        23.6 %     ( 22 %)     Nodeslider(V)
         NA           NA       TLMultiplier(V)

  Chain swap information for run 1:

             1     2     3     4 
       --------------------------
     1 |        0.03  0.01  0.00 
     2 |   178        0.16  0.05 
     3 |   169   149        0.31 
     4 |   166   177   161       

  Chain swap information for run 2:

             1     2     3     4 
       --------------------------
     1 |        0.19  0.04  0.01 
     2 |   162        0.39  0.02 
     3 |   190   152        0.15 
     4 |   177   170   149       

  Upper diagonal: Proportion of successful state exchanges between chains
  Lower diagonal: Number of attempted state exchanges between chains

  Chain information:

    ID -- Heat 
   -----------
     1 -- 1.00  (cold chain)
     2 -- 0.91 
     3 -- 0.83 
     4 -- 0.77 

  Heat = 1 / (1 + T * (ID - 1))
     (where T = 0.10 is the temperature and ID is the chain number)

  Reached end of file

Tasks completed, exiting program because mode is noninteractive To return control to the command line after completion of file processing, set mode to interactive with 'mb -i ' (i is for interactive) or use 'set mode=interactive'

   1000 -- (-6102.739) (-5942.009) [-5903.369] (-6034.000) * [-5909.130] (-6047.537) (-5973.908) (-5969.691) -- 0:00:00

  Average standard deviation of split frequencies: 0.006203
  Final log likelihoods and log prior probs for run 1 (stored and calculated):
     Chain 1 -- -6102.739461 -- 15.229932
     Chain 1 -- -6102.739480 -- 15.229932
     Chain 2 -- -5942.009386 -- 14.523801
     Chain 2 -- -5942.009363 -- 14.523801
     Chain 3 -- -5903.369331 -- 13.945189
     Chain 3 -- -5903.369429 -- 13.945189
     Chain 4 -- -6033.999928 -- 14.421609
     Chain 4 -- -6033.999872 -- 14.421609
  Final log likelihoods and log prior probs for run 2 (stored and calculated):
     Chain 1 -- -5909.129855 -- 13.061768
     Chain 1 -- -5909.129829 -- 13.061768
     Chain 2 -- -6047.537238 -- 13.588080
     Chain 2 -- -6047.537238 -- 13.588080
     Chain 3 -- -5973.908051 -- 14.274808
     Chain 3 -- -5973.907985 -- 14.274808
     Chain 4 -- -5969.690943 -- 15.615901
     Chain 4 -- -5969.690943 -- 15.615901

  Analysis completed in 1 second
  Analysis used 1.43 seconds of CPU time
  Likelihood of best state for "cold" chain of run 1 was -5900.63
  Likelihood of best state for "cold" chain of run 2 was -5941.80

  Acceptance rates for the moves in the "cold" chain of run 1:
     With prob.   (last 100)   chain accepted proposals by move
         NA           NA       Dirichlet(Revmat)
         NA           NA       Slider(Revmat)
         NA           NA       Dirichlet(Pi)
         NA           NA       Slider(Pi)
         NA           NA       Multiplier(Alpha)
         NA           NA       Slider(Pinvar)
         NA           NA       ExtSPR(Tau,V)
         NA           NA       ExtTBR(Tau,V)
         0.9 %     (  1 %)     NNI(Tau,V)
         6.3 %     (  7 %)     ParsSPR(Tau,V)
        44.2 %     ( 38 %)     Multiplier(V)
        26.0 %     ( 25 %)     Nodeslider(V)
         NA           NA       TLMultiplier(V)

  Acceptance rates for the moves in the "cold" chain of run 2:
     With prob.   (last 100)   chain accepted proposals by move
         NA           NA       Dirichlet(Revmat)
         NA           NA       Slider(Revmat)
         NA           NA       Dirichlet(Pi)
         NA           NA       Slider(Pi)
         NA           NA       Multiplier(Alpha)
         NA           NA       Slider(Pinvar)
         4.2 %     (  5 %)     ExtSPR(Tau,V)
         1.7 %     (  2 %)     ExtTBR(Tau,V)
         4.5 %     (  5 %)     NNI(Tau,V)
         8.8 %     (  9 %)     ParsSPR(Tau,V)
        48.1 %     ( 42 %)     Multiplier(V)
        23.6 %     ( 22 %)     Nodeslider(V)
         NA           NA       TLMultiplier(V)

  Chain swap information for run 1:

             1     2     3     4 
       --------------------------
     1 |        0.03  0.01  0.00 
     2 |   178        0.16  0.05 
     3 |   169   149        0.31 
     4 |   166   177   161       

  Chain swap information for run 2:

             1     2     3     4 
       --------------------------
     1 |        0.19  0.04  0.01 
     2 |   162        0.39  0.02 
     3 |   190   152        0.15 
     4 |   177   170   149       

  Upper diagonal: Proportion of successful state exchanges between chains
  Lower diagonal: Number of attempted state exchanges between chains

  Chain information:

    ID -- Heat 
   -----------
     1 -- 1.00  (cold chain)
     2 -- 0.91 
     3 -- 0.83 
     4 -- 0.77 

  Heat = 1 / (1 + T * (ID - 1))
     (where T = 0.10 is the temperature and ID is the chain number)

  Reached end of file

Tasks completed, exiting program because mode is noninteractive To return control to the command line after completion of file processing, set mode to interactive with 'mb -i ' (i is for interactive) or use 'set mode=interactive'

   1000 -- (-6102.739) (-5942.009) [-5903.369] (-6034.000) * [-5909.130] (-6047.537) (-5973.908) (-5969.691) -- 0:00:00

  Average standard deviation of split frequencies: 0.006203
  Final log likelihoods and log prior probs for run 1 (stored and calculated):
     Chain 1 -- -6102.739461 -- 15.229932
     Chain 1 -- -6102.739480 -- 15.229932
     Chain 2 -- -5942.009386 -- 14.523801
     Chain 2 -- -5942.009363 -- 14.523801
     Chain 3 -- -5903.369331 -- 13.945189
     Chain 3 -- -5903.369429 -- 13.945189
     Chain 4 -- -6033.999928 -- 14.421609
     Chain 4 -- -6033.999872 -- 14.421609
  Final log likelihoods and log prior probs for run 2 (stored and calculated):
     Chain 1 -- -5909.129855 -- 13.061768
     Chain 1 -- -5909.129829 -- 13.061768
     Chain 2 -- -6047.537238 -- 13.588080
     Chain 2 -- -6047.537238 -- 13.588080
     Chain 3 -- -5973.908051 -- 14.274808
     Chain 3 -- -5973.907985 -- 14.274808
     Chain 4 -- -5969.690943 -- 15.615901
     Chain 4 -- -5969.690943 -- 15.615901

  Analysis completed in 1 second
  Analysis used 1.44 seconds of CPU time
  Likelihood of best state for "cold" chain of run 1 was -5900.63
  Likelihood of best state for "cold" chain of run 2 was -5941.80

  Acceptance rates for the moves in the "cold" chain of run 1:
     With prob.   (last 100)   chain accepted proposals by move
         NA           NA       Dirichlet(Revmat)
         NA           NA       Slider(Revmat)
         NA           NA       Dirichlet(Pi)
         NA           NA       Slider(Pi)
         NA           NA       Multiplier(Alpha)
         NA           NA       Slider(Pinvar)
         NA           NA       ExtSPR(Tau,V)
         NA           NA       ExtTBR(Tau,V)
         0.9 %     (  1 %)     NNI(Tau,V)
         6.3 %     (  7 %)     ParsSPR(Tau,V)
        44.2 %     ( 38 %)     Multiplier(V)
        26.0 %     ( 25 %)     Nodeslider(V)
         NA           NA       TLMultiplier(V)

  Acceptance rates for the moves in the "cold" chain of run 2:
     With prob.   (last 100)   chain accepted proposals by move
         NA           NA       Dirichlet(Revmat)
         NA           NA       Slider(Revmat)
         NA           NA       Dirichlet(Pi)
         NA           NA       Slider(Pi)
         NA           NA       Multiplier(Alpha)
         NA           NA       Slider(Pinvar)
         4.2 %     (  5 %)     ExtSPR(Tau,V)
         1.7 %     (  2 %)     ExtTBR(Tau,V)
         4.5 %     (  5 %)     NNI(Tau,V)
         8.8 %     (  9 %)     ParsSPR(Tau,V)
        48.1 %     ( 42 %)     Multiplier(V)
        23.6 %     ( 22 %)     Nodeslider(V)
         NA           NA       TLMultiplier(V)

  Chain swap information for run 1:

             1     2     3     4 
       --------------------------
     1 |        0.03  0.01  0.00 
     2 |   178        0.16  0.05 
     3 |   169   149        0.31 
     4 |   166   177   161       

  Chain swap information for run 2:

             1     2     3     4 
       --------------------------
     1 |        0.19  0.04  0.01 
     2 |   162        0.39  0.02 
     3 |   190   152        0.15 
     4 |   177   170   149       

  Upper diagonal: Proportion of successful state exchanges between chains
  Lower diagonal: Number of attempted state exchanges between chains

  Chain information:

    ID -- Heat 
   -----------
     1 -- 1.00  (cold chain)
     2 -- 0.91 
     3 -- 0.83 
     4 -- 0.77 

  Heat = 1 / (1 + T * (ID - 1))
     (where T = 0.10 is the temperature and ID is the chain number)

  Reached end of file

Tasks completed, exiting program because mode is noninteractive To return control to the command line after completion of file processing, set mode to interactive with 'mb -i ' (i is for interactive) or use 'set mode=interactive'

zhangchicool commented 6 years ago

Just by looking at the header of the output, the serial version is launched, not the parallel version. The parallel version should show something like this: MrBayes 3.2.6 x86_64

                  (Bayesian Analysis of Phylogeny)

                         (Parallel version)
                     (8 processors available)
ferroao commented 6 years ago

I will go for revBayes

nylander commented 5 years ago

User apparently did not use the mpi version of "mb", but the serial, when using mpirun.

sun-jiao commented 2 years ago

User apparently did not use the mpi version of "mb", but the serial, when using mpirun.

I'm very sure it's the parallel version but facing the same problem:

image

mpirun ran it 8 times using 1 processor.

kusalananda commented 2 years ago

User apparently did not use the mpi version of "mb", but the serial, when using mpirun.

The original user was using the serial version of MrBayes (but had the parallel version installed elsewhere).

I'm very sure it's the parallel version but facing the same problem:

image

mpirun ran it 8 times using 1 processor.

It's unclear how you started MrBayes here and why you are using it in interactive mode.

sun-jiao commented 2 years ago

@kusalananda so it's not right to use parallel version in interactive mode?

kusalananda commented 2 years ago

@kusalananda so it's not right to use parallel version in interactive mode?

It makes very little sense to use the parallel version in interactive mode. When running MrBayes in parallel with MPI, you are expected to provide it with an input file instead of interactive commands. Many users of the MPI variant of MrBayes would run it on a shared compute cluster by submitting a non-interactive job to a job queuing system, together with some job resource description (depending on the queuing system).

One could use the non-parallel variant of MrBayes to try things out and run smaller trial runs before starting a more extensive run on multiple CPUs and possibly numerous connected systems.

There would additionally not be any benefit in running the parallel variant of MrBayes on a uniprocessor system.

sun-jiao commented 2 years ago

@kusalananda so it's not right to use parallel version in interactive mode?

It makes very little sense to use the parallel version in interactive mode. When running MrBayes in parallel with MPI, you are expected to provide it with an input file instead of interactive commands. Many users of the MPI variant of MrBayes would run it on a shared compute cluster by submitting a non-interactive job to a job queuing system, together with some job resource description (depending on the queuing system).

One could use the non-parallel variant of MrBayes to try things out and run smaller trial runs before starting a more extensive run on multiple CPUs and possibly numerous connected systems.

There would additionally not be any benefit in running the parallel variant of MrBayes on a uniprocessor system.

Thanks, I changed from mpich to openmpi, which fixed the above problem. The speed increased a bit in my laptop (estimated time from 66 hours decreased to 36 hours).

  • © Githubissues.
  • Githubissues is a development platform for aggregating issues.