def get_model_difficulty(modelstring):
"""
Input a model string like HKY+I+G or LG+G+F, and a guess about how long it
takes to analyse Right now, this is done with a simple hack. I just return
a number that is the number of params plus a modifier for extra stuff like
+I and +G the hardest models are +I+G, then +G, then +I this is just used
to rank models for ordering the analysis. The return is a 'difficulty'
score that can be used to rank models for queing and do the hardest first
"""
elements = modelstring.split("+")
model_params = get_num_params(modelstring)
difficulty = 0
if "G" in elements[1:]:
difficulty += 2000
if "I" in elements[1:]:
difficulty += 1000
if "F" in elements[1:]:
difficulty += 3000
if "X" in elements[1:]:
difficulty += 4000
if the_config.datatype == "protein" and "GTR" in modelstring:
# that's a tough model with 189 free parameters
difficulty += 10000
extras = modelstring.count("+")
total = model_params + extras + difficulty
log.debug("Model: %s Difficulty: %d" % (modelstring, total))
return total
def get_model_difficulty(modelstring):
"""
Input a model string like HKY+I+G or LG+G+F, and a guess about how long it
takes to analyse Right now, this is done with a simple hack. I just return
a number that is the number of params plus a modifier for extra stuff like
+I and +G the hardest models are +I+G, then +G, then +I this is just used
to rank models for ordering the analysis. The return is a 'difficulty'
score that can be used to rank models for queing and do the hardest first
"""
elements = modelstring.split("+")
model_params = get_num_params(modelstring)
difficulty = 0
if "G" in elements[1:]:
difficulty += 2000
if "I" in elements[1:]:
difficulty += 1000
if "F" in elements[1:]:
difficulty += 3000
if "X" in elements[1:]:
difficulty += 4000
if the_config.datatype == "protein" and "GTR" in modelstring:
# that's a tough model with 189 free parameters
difficulty += 10000
extras = modelstring.count("+")
total = model_params + extras + difficulty
log.debug("Model: %s Difficulty: %d" % (modelstring, total))
return total
if "LG4" in modelstring:
# these models are hard
difficulty += 9000
extras = modelstring.count("+")
total = model_params + extras + difficulty
log.debug("Model: %s Difficulty: %d" % (modelstring, total))
return total
if __name__ == "__main__":
print " ",
print "Name".ljust(15),
print "Params".ljust(10),
print "Diff".ljust(10),
print "CommandLine"
for i, model in enumerate(get_all_dna_models()):
print str(i + 1).rjust(2),
print model.ljust(15),
print str(get_num_params(model)).ljust(10),
print str(get_model_difficulty(model)).ljust(10),
print get_model_commandline(model)
for i, model in enumerate(get_all_protein_models()):
print str(i + 1).rjust(2),
print model.ljust(15),
print str(get_num_params(model)).ljust(10),
print str(get_model_difficulty(model)).ljust(10),
print get_model_commandline(model)
if the_config.datatype == "protein" and "GTR" in modelstring:
# that's a tough model with 189 free parameters
difficulty += 10000
extras = modelstring.count("+")
total = model_params + extras + difficulty
log.debug("Model: %s Difficulty: %d" % (modelstring, total))
return total
if __name__ == "__main__":
print " ",
print "Name".ljust(15),
print "Params".ljust(10),
print "Diff".ljust(10),
print "CommandLine"
for i, model in enumerate(get_all_dna_models()):
print str(i+1).rjust(2),
print model.ljust(15),
print str(get_num_params(model)).ljust(10),
print str(get_model_difficulty(model)).ljust(10),
print get_model_commandline(model)
for i, model in enumerate(get_all_protein_models()):
print str(i+1).rjust(2),
print model.ljust(15),
print str(get_num_params(model)).ljust(10),
print str(get_model_difficulty(model)).ljust(10),
print get_model_commandline(model)
