search

radical-cybertools / radical.repex.at

This is the github location for RepEx developed by the RADICAL team in conjunction with the York Lab.
Other
4 stars 3 forks source link

TSU problem in perfopt_gen branch #78

Closed haoyuanchen closed 8 years ago

haoyuanchen commented 8 years ago

The simulation cannot run because in the mdin files, the @salt@ placeholder were not replaced by corresponding values.

antonst commented 8 years ago

Can you please paste you input file here?

haoyuanchen commented 8 years ago

JSON file:

{ "remd.input": { "re_pattern": "S", "number_of_cycles": "3", "input_folder": "tsu_remd_inputs", "input_file_basename": "dna_remd", "amber_input": "dna.mdin", "us_template": "dna_us.RST", "amber_parameters": "dna.parm7", "amber_coordinates_folder": "dna_coors", "same_coordinates": "False", "group_exec": "False", "init_temp": "288.0", "replica_mpi": "False", "replica_cores": "1", "steps_per_cycle": "1000", "download_mdinfo": "False", "download_mdout" : "False" }, "dim.input": { "d2": { "type" : "salt", "number_of_replicas": "4", "exchange_replica_cores" : "4", "min_salt": "0.0", "max_salt": "0.15" }, "d1": { "type" : "temperature", "number_of_replicas": "2", "min_temperature": "288.0", "max_temperature": "289.0" }, "d3": { "type" : "umbrella", "number_of_replicas": "12", "min_us_param": "67.5", "max_us_param": "97.5" } } }

Template mdin file:

NVT equilibration &cntrl imin = 0, ntx = @ntx@, irest = @irest@, ntpr = @nstlim@, ntwr = @nstlim@, ntwx = @nstlim@, ntf = 2, ntc = 2, ntb = 0, igb = 1, cut = 999.0, nstlim = @nstlim@, dt = 0.002, extdiel = 81.946, saltcon = @salt@, temp0 = @temp@, ntt = 3, gamma_ln = 5.0, ig = -1, nmropt = 1 / &wt type = 'DUMPFREQ', istep1 = 10, / &wt type = 'END' / DISANG = @disang@ DUMPAVE = @disang@.out

antonst commented 8 years ago

There is no dna_remd in examples directory, so I can't test your input file, but I have pushed a fix which should solve this problem. Can you please try to re-run your simulation?

antonst commented 8 years ago

Sorry, I have found input folder. It seems to work fine, I just had to rename coordinate files, so that they don't have a .0 at the end.

antonst commented 8 years ago

Can you please verify this?

haoyuanchen commented 8 years ago

Now the simulation could run. I'm seeing exchanges in both T and S dimensions, but not U dimension. That might be related to the exchange issue we've been discussing, but we can close this ticket now.

Thanks!