def generate_and_compare(self, name):
directory = os.path.dirname(__file__)
name = os.path.join(directory, name)
cds = CaseDataset(name + '.json', 'json')
data = cds.data.fetch()
caseset_query_to_csv(data, self.filename_csv)
with open(name + '.csv', 'r') as inp1:
expected = inp1.readlines()
with open(self.filename_csv, 'r') as inp2:
actual = inp2.readlines()
# Strip off trailing whitespace (newlines and carriage returns)
# Don't check time-stamp because some OS round it.
for exp, act in zip(expected, actual):
# skip timestamps, and uuids
items2 = act.rstrip().split(",")[1:-3]
for i, item1 in enumerate(exp.rstrip().split(",")[1:-3]):
item2 = items2[i]
try: # (str).isnumeric() only works on unicode
item1, item2 = float(item1), float(item2)
# nan equality check fails by definition
if isnan(item1) and isnan(item2):
continue
self.assertEqual(item1, item2)
except (ValueError, TypeError):
self.assertEqual(item1, item2)
def generate_and_compare(self, name):
directory = os.path.dirname(__file__)
name = os.path.join(directory, name)
cds = CaseDataset(name + '.json', 'json')
data = cds.data.fetch()
caseset_query_to_csv(data, self.filename_csv)
with open(name + '.csv', 'r') as inp1:
expected = inp1.readlines()
with open(self.filename_csv, 'r') as inp2:
actual = inp2.readlines()
# Strip off trailing whitespace (newlines and carriage returns)
# Don't check time-stamp because some OS round it.
for exp, act in zip(expected, actual):
# skip timestamps, and uuids
items2 = act.rstrip().split(",")[1:-3]
for i, item1 in enumerate(exp.rstrip().split(",")[1:-3]):
item2 = items2[i]
try: # (str).isnumeric() only works on unicode
item1, item2 = float(item1), float(item2)
# nan equality check fails by definition
if isnan(item1) and isnan(item2):
continue
self.assertEqual(item1, item2)
except (ValueError, TypeError):
self.assertEqual(item1, item2)
def test_simple(self):
# Make sure the CSV file can be read and has the correct number of cases
self.top.recorders = [JSONCaseRecorder(self.filename_json)]
self.top.recorders[0].num_backups = 0
self.top.run()
cds = CaseDataset(self.filename_json, 'json')
data = cds.data.fetch() # results
caseset_query_to_csv(data, self.filename_csv)
cases = [case for case in CSVCaseIterator(filename=self.filename_csv)]
def test_simple(self):
# Make sure the CSV file can be read and has the correct number of cases
self.top.recorders = [JSONCaseRecorder(self.filename_json)]
self.top.recorders[0].num_backups = 0
self.top.run()
cds = CaseDataset(self.filename_json, 'json')
data = cds.data.fetch() # results
caseset_query_to_csv( data, self.filename_csv)
cases = [case for case in CSVCaseIterator(filename=self.filename_csv)]
self.assertEqual(len(cases), 10)
def test_flatten(self):
# try it after creating some Cases
# more rigorous checking of the csv
outputs = ['comp1.a_array', 'comp1.vt']
inputs = [('comp1.x_array', array([2.0, 2.0, 2.0]))]
cases = [Case(inputs=inputs, outputs=outputs)]
self.top.driver.clear_parameters()
Case.set_vartree_inputs(self.top.driver, cases)
self.top.driver.clear_responses()
self.top.driver.add_responses(outputs)
self.top.recorders = [JSONCaseRecorder(self.filename_json)]
self.top.recorders[0].num_backups = 0
self.top.run()
cds = CaseDataset(self.filename_json, 'json')
data = cds.data.fetch() # results
caseset_query_to_csv(data, self.filename_csv)
# check recorded cases
cases = [case for case in CSVCaseIterator(filename=self.filename_csv)]
def test_flatten(self):
# try it after creating some Cases
# more rigorous checking of the csv
outputs = ['comp1.a_array', 'comp1.vt']
inputs = [('comp1.x_array', array([2.0, 2.0, 2.0]))]
cases = [Case(inputs=inputs, outputs=outputs)]
self.top.driver.clear_parameters()
Case.set_vartree_inputs(self.top.driver, cases)
self.top.driver.clear_responses()
self.top.driver.add_responses(outputs)
self.top.recorders = [JSONCaseRecorder(self.filename_json)]
self.top.recorders[0].num_backups = 0
self.top.run()
cds = CaseDataset(self.filename_json, 'json')
data = cds.data.fetch() # results
caseset_query_to_csv( data, self.filename_csv)
# check recorded cases
cases = [case for case in CSVCaseIterator(filename=self.filename_csv)]
sout = StringIO.StringIO()
for case in cases:
print >>sout, case
expected = [
'Case:',
' uuid: ad4c1b76-64fb-11e0-95a8-001e8cf75fe',
' timestamp: 1383238593.781986',
' parent_uuid: ad4c1b76-64fb-11e0-95a8-001e8cf75fe',
' inputs:',
' comp1.x_array[0]: 2.0',
' comp1.x_array[1]: 2.0',
' comp1.x_array[2]: 2.0',
' outputs:',
' Response(comp1.a_array)[0]: 1.0',
' Response(comp1.a_array)[1]: 3.0',
' Response(comp1.a_array)[2]: 5.5',
' Response(comp1.vt).data: ',
' Response(comp1.vt).v1: 1.0',
' Response(comp1.vt).v2: 2.0',
' Response(comp1.vt).vt2.data: ',
' Response(comp1.vt).vt2.vt3.a: 1.0',
' Response(comp1.vt).vt2.vt3.b: 12.0',
' Response(comp1.vt).vt2.vt3.data: ',
' Response(comp1.vt).vt2.x: -1.0',
' Response(comp1.vt).vt2.y: -2.0',
' comp1.a_array[0]: 1.0',
' comp1.a_array[1]: 3.0',
' comp1.a_array[2]: 5.5',
" comp1.a_string: Hello',;','",
' comp1.derivative_exec_count: 0.0',
' comp1.exec_count: 1.0',
' comp1.itername: 1-comp1',
' comp1.vt.data: ',
' comp1.vt.v1: 1.0',
' comp1.vt.v2: 2.0',
' comp1.vt.vt2.data: ',
' comp1.vt.vt2.vt3.a: 1.0',
' comp1.vt.vt2.vt3.b: 12.0',
' comp1.vt.vt2.vt3.data: ',
' comp1.vt.vt2.x: -1.0',
' comp1.vt.vt2.y: -2.0',
' comp1.z: 0.0',
' comp2.derivative_exec_count: 0.0',
' comp2.exec_count: 1.0',
' comp2.itername: 1-comp2',
' comp2.z: 1.0',
' driver.workflow.itername: 1',
]
#print sout.getvalue()
lines = sout.getvalue().split('\n')
for index, line in enumerate(lines):
if line.startswith('Case:'):
for i in range(len(expected)):
if expected[i].startswith(' uuid:'):
self.assertTrue(lines[index+i].startswith(' uuid:'))
elif expected[i].startswith(' parent_uuid:'):
self.assertTrue(lines[index+i].startswith(' parent_uuid:'))
elif expected[i].startswith(' timestamp:'):
self.assertTrue(lines[index+i].startswith(' timestamp:'))
else:
self.assertEqual(lines[index+i], expected[i])
break
else:
self.fail("couldn't find the expected Case")
def test_flatten(self):
# try it after creating some Cases
# more rigorous checking of the csv
outputs = ['comp1.a_array', 'comp1.vt']
inputs = [('comp1.x_array', array([2.0, 2.0, 2.0]))]
cases = [Case(inputs=inputs, outputs=outputs)]
self.top.driver.clear_parameters()
Case.set_vartree_inputs(self.top.driver, cases)
self.top.driver.clear_responses()
self.top.driver.add_responses(outputs)
self.top.recorders = [JSONCaseRecorder(self.filename_json)]
self.top.recorders[0].num_backups = 0
self.top.run()
cds = CaseDataset(self.filename_json, 'json')
data = cds.data.fetch() # results
caseset_query_to_csv( data, self.filename_csv)
# check recorded cases
cases = [case for case in CSVCaseIterator(filename=self.filename_csv)]
sout = StringIO.StringIO()
for case in cases:
print >>sout, case
expected = \
'''Case:
uuid: 07280785-9b76-11e4-800d-20c9d0478eff
timestamp: 1421189195.646824
parent_uuid: 0720c385-9b76-11e4-b796-20c9d0478eff
inputs:
comp1.x_array[0]: 2.0
comp1.x_array[1]: 2.0
comp1.x_array[2]: 2.0
outputs:
_pseudo_4.out0[0]: 1.0
_pseudo_4.out0[1]: 3.0
_pseudo_4.out0[2]: 5.5
_pseudo_5.out0.data:
_pseudo_5.out0.v1: 1.0
_pseudo_5.out0.v2: 2.0
_pseudo_5.out0.vt2.data:
_pseudo_5.out0.vt2.vt3.a: 1.0
_pseudo_5.out0.vt2.vt3.b: 12.0
_pseudo_5.out0.vt2.vt3.data:
_pseudo_5.out0.vt2.x: -1.0
_pseudo_5.out0.vt2.y: -2.0
comp1.a_array[0]: 1.0
comp1.a_array[1]: 3.0
comp1.a_array[2]: 5.5
comp1.a_string: Hello',;','
comp1.derivative_exec_count: 0.0
comp1.exec_count: 1.0
comp1.itername: 1-comp1
comp1.vt.data:
comp1.vt.v1: 1.0
comp1.vt.v2: 2.0
comp1.vt.vt2.data:
comp1.vt.vt2.vt3.a: 1.0
comp1.vt.vt2.vt3.b: 12.0
comp1.vt.vt2.vt3.data:
comp1.vt.vt2.x: -1.0
comp1.vt.vt2.y: -2.0
comp1.z: 0.0
comp2.derivative_exec_count: 0.0
comp2.exec_count: 1.0
comp2.itername: 1-comp2
comp2.z: 1.0
driver.workflow.itername: 1
'''
print sout.getvalue()
lines = sout.getvalue().split('\n')
expected_lines = expected.split('\n')
for index, line in enumerate(lines):
if line.startswith('Case:'):
for i in range(len(expected_lines)):
if expected_lines[i].startswith(' uuid:'):
self.assertTrue(lines[index+i].startswith(' uuid:'))
elif expected_lines[i].startswith(' parent_uuid:'):
self.assertTrue(lines[index+i].startswith(' parent_uuid:'))
elif expected_lines[i].startswith(' timestamp:'):
self.assertTrue(lines[index+i].startswith(' timestamp:'))
else:
self.assertEqual(lines[index+i], expected_lines[i])
break
else:
self.fail("couldn't find the expected Case")
self.connect('supin.Outputs.M2', ['cart3d.M2', 'fun3d.M2'])
self.connect('supin.Outputs.A2', ['cart3d.A2', 'fun3d.A2'])
self.connect('supin.Outputs.mdot', ['cart3d.mdot', 'fun3d.mdot'])
self.connect('supin.Outputs.yEF', ['opencsm.yEF'])
self.connect('supin.Outputs.Rspin', ['opencsm.Rspin'])
self.recorders = [JSONCaseRecorder(out=OPT_DB)]
if __name__ == '__main__':
if os.path.exists(OPT_DB):
os.remove(OPT_DB)
if os.path.exists(OPT_CSV):
os.remove(OPT_CSV)
OPT = Analysis()
# --------------------------------------------------------------------------- #
# --- Execute OpenMDAO Analysis ---
# --------------------------------------------------------------------------- #
OPT.run()
#----------------------------------------------------
# Print out history of our objective for inspection
#----------------------------------------------------
case_dataset = CaseDataset(OPT_DB, 'json')
data = case_dataset.data.by_case().fetch()
caseset_query_to_csv(data, filename='doe.csv')
