def _testIterator(self, itr):
for item in itr:
model = item.model
self.assertTrue(isinstance(model.getSpecies(0), libsbml.Species))
COUNT = 20
itr = simple_sbml.modelIterator(final=COUNT)
item_number = -1
for item in itr:
self.assertTrue(isinstance(item.filename, str))
self.assertTrue(util.isSBMLModel(item.model))
item_number = item.number
self.assertEqual(item_number, COUNT - 1)
def calcStats(initial=0, final=50, out_path=OUTPUT_PATH,
report_interval=50, report_progress=True, min_frc=-1,
data_dir=cn.BIOMODELS_DIR):
"""
Calculates statistics for structured names.
:param int initial: Index of first model to process
:param int final: Index of final model to process
:param str out_path: Path to the output CSV file
:param int report_interval: Number of files processed before
a report is written
:param bool report_progress: report file being processed
:param float min_frc: Filter to select only those models
that have at least the specified fraction of reactions
balanced according to moiety_analysis
"""
def writeDF(dfs):
df_count = pd.concat(dfs)
df_count[cn.NUM_BALANCED_REACTIONS] = \
df_count[cn.TOTAL_REACTIONS] \
- df_count[cn.NUM_IMBALANCED_REACTIONS]
denom = (df_count[cn.TOTAL_REACTIONS]
- df_count[cn.NUM_BOUNDARY_REACTIONS])
denom = [np.nan if np.isclose(v, 0) else v for v in denom]
df_count[cn.FRAC_BALANCED_REACTIONS] = \
1.0*df_count[cn.NUM_BALANCED_REACTIONS] / denom
df_count[cn.FRAC_BOUNDARY_REACTIONS] = \
1.0*df_count[cn.NUM_BOUNDARY_REACTIONS] / (
df_count[cn.TOTAL_REACTIONS])
if min_frc < 0:
df = df_count
else:
df = df_count[df_count[cn.FRAC_BALANCED_REACTIONS] > min_frc]
df = df.sort_values(cn.FRAC_BALANCED_REACTIONS)
df.to_csv(out_path, index=False)
#
dfs = []
sbmliter = simple_sbml.modelIterator(initial=initial, final=final,
data_dir=data_dir)
for item in sbmliter:
if report_progress:
print("*Processing file %s, number %d"
% (item.filename, item.number))
simple = simple_sbml.SimpleSBML()
try:
simple.initialize(item.model)
except:
print(" Error in model number %d." % item.number)
continue
row = {cn.FILENAME: [item.filename],
cn.IS_STRUCTURED: [False],
cn.NUM_BOUNDARY_REACTIONS: [0],
cn.TOTAL_REACTIONS: [0],
cn.NUM_IMBALANCED_REACTIONS: [0],
}
for reaction in simple.reactions:
if (len(reaction.reactants) == 0) or (len(reaction.products) == 0):
row[cn.NUM_BOUNDARY_REACTIONS] = \
[row[cn.NUM_BOUNDARY_REACTIONS][0] + 1]
molecules = util.uniqueify([m.molecule
for m in set(reaction.reactants).union(reaction.products)])
if any([isStructuredName(m.name) for m in molecules]):
row[cn.IS_STRUCTURED] = [True]
try:
mcr = sbmllint.lint(model_reference=item.model, is_report=False)
row[cn.TOTAL_REACTIONS] = [mcr.num_reactions if mcr.num_reactions > 0 else np.nan]
row[cn.NUM_IMBALANCED_REACTIONS] = [mcr.num_imbalances]
except:
row[cn.TOTAL_REACTIONS] = [None]
row[cn.NUM_IMBALANCED_REACTIONS] = [0]
dfs.append(pd.DataFrame(row))
if item.number % report_interval == 0:
writeDF(dfs)
writeDF(dfs)
def testModelIterator2(self):
if IGNORE_TEST:
return
self._testIterator(
simple_sbml.modelIterator(final=1, zip_filename=None))
def testModelIterator1(self):
if IGNORE_TEST:
return
self._testIterator(simple_sbml.modelIterator(final=1))