def generate_data(model, sim_length, inputdir, outputdir):
"""
The first pipeline step: data generation.
:param model: the model to process
:param sim_length: the length of the simulation
:param inputdir: the directory containing the model
:param outputdir: the directory to store the results
:return: no output
"""
if not os.path.isfile(os.path.join(inputdir, model)):
logger.error(os.path.join(inputdir, model) + " does not exist.")
return
refresh_directory(outputdir, model[:-4])
logger.info("Simulating Model: " + model)
model_noext = model[:-4]
copasi = get_copasi()
if copasi is None:
logger.error("CopasiSE not found! Please check that CopasiSE is "
"installed and in the PATH environmental variable.")
return
# run CopasiSE. Copasi must generate a (TIME COURSE) report
process = subprocess.Popen([copasi, '--nologo', os.path.join(inputdir, model)])
process.wait()
if (not os.path.isfile(os.path.join(inputdir, model_noext + ".csv")) and
not os.path.isfile(os.path.join(inputdir, model_noext + ".txt"))):
logger.warn(os.path.join(inputdir, model_noext + ".csv") + " (or .txt) does not exist!")
return
if os.path.isfile(os.path.join(inputdir, model_noext + ".txt")):
os.rename(os.path.join(inputdir, model_noext + ".txt"), os.path.join(inputdir, model_noext + ".csv"))
# Replace some string in the report file
replace_str_copasi_sim_report(os.path.join(inputdir, model_noext + ".csv"))
# copy file removing empty lines
with open(os.path.join(inputdir, model_noext + ".csv"), 'r') as filein, \
open(os.path.join(outputdir, model_noext + ".csv"), 'w') as fileout:
for line in filein:
if not line.isspace():
fileout.write(line)
os.remove(os.path.join(inputdir, model_noext + ".csv"))
# Extract a selected time point from all perturbed time courses contained in the report file
with open(os.path.join(outputdir, model_noext + ".csv"), 'r') as filein:
lines = filein.readlines()
header = lines[0]
lines = lines[1:]
timepoints = range(0, sim_length + 1)
filesout = []
try:
filesout = [open(os.path.join(outputdir, model_noext + "__tp_%d.csv" % i), "w") for i in timepoints]
# copy the header
for fileout in filesout:
fileout.write(header)
# extract the i-th time point and copy it to the corresponding i-th file
for line in lines:
tp = line.rstrip().split('\t')[0]
if not '.' in tp and int(tp) in timepoints:
filesout[int(tp)].write(line)
finally:
for fileout in filesout:
fileout.close()
def generate_data(model, inputdir, outputdir, cluster_type="pp", pp_cpus=2, runs=1):
"""
The first pipeline step: data generation.
:param model: the model to process
:param inputdir: the directory containing the model
:param outputdir: the directory containing the output files
:param cluster_type: pp for local Parallel Python, lsf for Load Sharing Facility, sge for Sun Grid Engine.
:param pp_cpus: the number of CPU used by Parallel Python.
:param runs: the number of model simulation
:return: no output
"""
if runs < 1:
logger.error("variable " + str(runs) + " must be greater than 0. Please, check your configuration file.")
return
if not os.path.isfile(os.path.join(inputdir, model)):
logger.error(os.path.join(inputdir, model) + " does not exist.")
return
copasi = get_copasi()
if copasi is None:
logger.error(
"CopasiSE not found! Please check that CopasiSE is installed and in the PATH environmental variable.")
return
# folder preparation
refresh_directory(outputdir, model[:-4])
# execute runs simulations.
logger.info("Simulating model " + model + " for " + str(runs) + " time(s)")
# Replicate the copasi file and rename its report file
groupid = "_" + get_rand_alphanum_str(20) + "_"
group_model = model[:-4] + groupid
for i in xrange(1, runs + 1):
shutil.copyfile(os.path.join(inputdir, model), os.path.join(inputdir, group_model) + str(i) + ".cps")
replace_string_in_file(os.path.join(inputdir, group_model) + str(i) + ".cps",
model[:-4] + ".csv",
group_model + str(i) + ".csv")
# run copasi in parallel
# To make things simple, the last 10 character of groupid are extracted and reversed.
# This string will be likely different from groupid and is the string to replace with
# the iteration number.
str_to_replace = groupid[10::-1]
command = copasi + " " + os.path.join(inputdir, group_model + str_to_replace + ".cps")
parallel_computation(command, str_to_replace, cluster_type, runs, outputdir, pp_cpus)
# move the report files
report_files = [f for f in os.listdir(inputdir) if
re.match(group_model + '[0-9]+.*\.csv', f) or re.match(group_model + '[0-9]+.*\.txt', f)]
for file in report_files:
# Replace some string in the report file
replace_str_copasi_sim_report(os.path.join(inputdir, file))
# rename and move the output file
shutil.move(os.path.join(inputdir, file), os.path.join(outputdir, file.replace(groupid, "_")[:-4] + ".csv"))
# removed repeated copasi files
repeated_copasi_files = [f for f in os.listdir(inputdir) if re.match(group_model + '[0-9]+.*\.cps', f)]
for file in repeated_copasi_files:
os.remove(os.path.join(inputdir, file))
