def toCCDCStructure(self, pointer=None, stage=None, identifier=None):
''' Returns CCDC structure instance from appropriate pointer
default atm is crystal optimiser
stage is SetCSPStructures.calculation = 'Data' (as general rule)
give identifier to change the name in the cif '''
from ccdc.io import CrystalReader
# temporary fail as code work in progress
if stage not in ['crystalOptimiserData', 'pcmData']:
raise Exception(
'Implement other options to get structures if needed')
#generalize this to be self.getAttribute(pointer = {stage: lowestStructurePointer}) if pointer is None
if pointer is None:
pointer = self.getAttribute({stage: 'finalStructurePointer'})
# CrystalReader can be told explicitly which file format to look for
if pointer[-3:].lower() in ['res',
'.16']: #.16 is fort.16 from DMACRYS
fileFormat = 'res'
elif pointer[-3:].lower() == 'cif':
fileFormat = 'cif'
ccdcCrystal = CrystalReader(pointer, format=fileFormat)[0]
if identifier:
ccdcCrystal.identifier = identifier
return ccdcCrystal
def main():
# names of files (all strings)
try:
try:
oldResFile, templateResFile, newFilename = sys.argv[1:]
except:
oldResFile, templateResFile, newFilename, enforcement = sys.argv[
1:]
except:
raise Exception(usefulMessage)
# use the above subroutines to achieve the relabelling
# uses ccdc molecule instances
with open(newFilename, 'w') as outf:
if len(sys.argv) == 4:
outf.write(
reorderedRes(
oldResFile,
replacementDict(
CrystalReader(templateResFile)[0],
CrystalReader(oldResFile)[0])))
elif len(sys.argv) == 5 and sys.argv[4] == 'y':
outf.write(
reorderedRes(oldResFile,
replacementDict(
CrystalReader(templateResFile)[0],
CrystalReader(oldResFile)[0]),
enforceStandardListOrder=True,
enforceLowestNumbering=True))
def cacl_density(doi):
print doi
query = TextNumericSearch()
query.add_doi(doi)
hits = query.search()
ref_codes = []
for hit in hits:
ref_codes.append(hit.identifier)
csd_crystal_reader = CrystalReader('CSD')
if len(ref_codes) == 0:
density = 1
else:
crystal = csd_crystal_reader.crystal(ref_codes[0])
density = crystal.calculated_density
return "%.3f" % density
def calc_powder_from_pdb(self, pdbfile):
crystal = CrystalReader(pdbfile)[0]
print crystal.spacegroup_symbol
print 'angles', crystal.cell_angles[0], crystal.cell_angles[1], crystal.cell_angles[2]
print 'length', crystal.cell_lengths[0], crystal.cell_lengths[1], crystal.cell_lengths[2]
print 'volume', crystal.cell_volume
pattern = PowderPattern.from_crystal(crystal)
pattern.write_xye_file(('./'+pdbfile.replace('.pdb','.xye')))
def test_ccdcToASE(self):
from ioAndInterfaces import ccdcCrystalToASE
from ccdc.io import CrystalReader
ccdcCrystal = CrystalReader('FPAMCA16.res')[0]
#print ccdcCrystal
aseCrystal = ccdcCrystalToASE(ccdcCrystal)
#print aseCrystal#.symbols
# this is a Z'=4 30 atom molecule
self.assertEqual(len(aseCrystal), 4 * 30)
def main():
try:
# oldString, oldResFile, templateResFile = sys.argv[1:]
oldString, resFile = sys.argv[1:]
except:
raise Exception(usefulMessage)
# print simpleRelabelling("+O1_C1_C2_C7 +H1_O1_C1_C2 +C1_C2_C7_N1 +C7_N1_C8_C9 +H6_N1_C8_C9 +F1_C14_C12_C11 +H6_N1_C8 +H1_O1_C1",
print simpleRelabelling("O1_C1_C2_C7 H1_O1_C1_C2 C1_C2_C7_N1 C7_N1_C8_C9 H6_N1_C8_C9 F1_C14_C12_C11 H6_N1_C8 H1_O1_C1",
CrystalReader(resFile)[0])
def _get_crystal(obj):
"""
Convert a obj's writable by MoleculeWriter to a crystal
:param `ccdc.molecule.Molecule` obj: molecule or protein
:return: `ccdc.crystal.Crystal`
"""
tmp = tempfile.mkdtemp()
f = os.path.join(tmp, "obj.mol2")
with MoleculeWriter(f) as w:
w.write(obj)
return CrystalReader(f)[0]
def test_SetCSPStructures(self):
from cspTracking import SetCSPStructures
tempSet = SetCSPStructures.initFromCrystalPredictorLog('exampleCP.log')
self.assertEqual(len(tempSet.structures), 27414)
self.assertEqual(tempSet.structures[0].crystalPredictorData['spaceGroup'], 'P-1')
self.assertEqual(tempSet.structures[5].idNumber, 6)
tempSet.structures[0].createCrystalOptimiserData('1-1/exampleCOSummary.out')
self.assertEqual(tempSet.structures[0].crystalOptimiserData['rhoFinal'],
1.4749)
#remove the data and add again with different method
tempSet.structures[0].crystalOptimiserData = {}
self.assertEqual(len(tempSet.structures[0].crystalOptimiserData.keys()), 0)
# tempSet.addCrystalOptimiserInfoFromGlobString('1-1/*ummary.out')
tempSet.addInfoFromGlobString('1-1/*ummary.out')
self.assertEqual(tempSet.structures[0].crystalOptimiserData['rhoFinal'],
1.4749)
self.assertTrue(tempSet.structures[0].crystalOptimiserComplete)
# get free energy data from Jonas's log files
tempSet.addFreeEnergyData(lambda x: 'free_energy.log')
self.assertEqual(tempSet.structures[-1].getAttribute({'freeEnergyData': 'ZPE'}),
2.7239)
#get a subset of data, and assert that it has a positive density
from cspTracking import Attribute
self.assertEqual(len(tempSet.subsetStructures([Attribute(pointer='crystalOptimiserComplete',
value = True)])),
len([x for x in tempSet.structures if x.crystalOptimiserComplete]))
self.assertEqual(len(tempSet.subsetStructures([Attribute(pointer = {'crystalOptimiserData': 'rhoFinal'},
minValue = 0.)])),
1)
self.assertEqual(len(tempSet.subsetStructures([Attribute(pointer = {'crystalOptimiserData': 'rhoFinal'},
maxValue = 0.)])),
0)
self.assertEqual(len(tempSet.subsetStructures([Attribute(pointer = {'crystalOptimiserData': 'rhoFinal'},
minValue = 0.1,
maxValue = 0.4)])),
0)
# tempSet.addCrystalOptimiserInfoFromGlobString('2-1/*ummary.out')
tempSet.addInfoFromGlobString('2-1/*ummary.out')
self.assertEqual(len(tempSet.subsetStructures([Attribute(pointer = {'crystalOptimiserData': 'rhoFinal'},
minValue = 1.1,
maxValue = 2.4)])),
2)
#lowest one for rho has idNumber 2
self.assertEqual([x.getAttribute('idNumber') for x in
tempSet.subsetStructures([Attribute(pointer = {'crystalOptimiserData': 'rhoFinal'},
minValue = 1.1,
maxValue = 2.4)],
sortAttribute=Attribute(pointer = {'crystalOptimiserData': 'rhoFinal'}))][0],
2)
#turn data into a pandas DataFrame
dataTable = tempSet.toPandasDataFrame()
self.assertEqual(dataTable['SG'][0], 'P-1')
#test matching
from ccdc.io import CrystalReader
from cspTracking import CSPStructure
matchingSet = SetCSPStructures([CSPStructure(1, initCrystalOptimiserData='1-1/exampleCOSummary.out')])
#match to itself just as an example
self.assertTrue(matchingSet.structures[0].crystalOptimiserData['finalStructureFileExists'])
matchDict = matchingSet.matchToStructureList([CrystalReader('1-1/lowest.res')[0]],
stage='crystalOptimiserData')
np.testing.assert_array_equal(matchDict['matchesMatrix'],
np.array([[15]]))
self.assertEqual(matchDict['candidateList'][0].idNumber,
1)
#do same match but add RMSD data
matchDict = matchingSet.matchToStructureList([CrystalReader('1-1/lowest.res')[0]],
returnRMSD=True,
stage='crystalOptimiserData')
np.testing.assert_array_equal(matchDict['rmsdMatrix'],
np.array([[0.]]))
#test packing coefficient (what method does CCDC use??)
self.assertAlmostEqual(matchingSet.structures[0].packingCoefficient(stage='crystalOptimiserData'),
0.69370122631)
#get minimum energy of list
self.assertEqual(matchingSet.minValueInSet(Attribute(pointer = {'crystalOptimiserData': 'eTotalFinal'})),
-133.44284)
#test writing a CIF - tear down data at the end
nameOfCIF = 'tempCIFWrite.cif'
self.assertFalse(os.path.isfile(nameOfCIF))
matchingSet.calculation = 'crystalOptimiser' # this should be known before matching really
matchingSet.makeCIF(nameOfCIF)
self.assertTrue(os.path.isfile(nameOfCIF))
os.remove(nameOfCIF)
#initiate via CSV then match
csvStructures = SetCSPStructures.initFromCSV('example.csv',
idNumberIndex = 0,
finalEnergyIndex = 1,
finalDensityIndex = 3,
finalStructureDir = os.path.dirname(os.path.realpath(__file__)),
assertZPrime = 1.234)
self.assertEqual(csvStructures.structures[0].idNumber,
978)
self.assertEqual(csvStructures.structures[0].getAttribute({'crystalOptimiserData': 'zPrime'}),
1.234)
self.assertAlmostEqual(csvStructures.structures[0].getAttribute({'crystalOptimiserData': 'eTotalFinal'}),
-139.35473)
self.assertAlmostEqual(csvStructures.structures[0].getAttribute({'crystalOptimiserData': 'rhoFinal'}),
1.26410)
print 'check match'
#use general get and set attributes
newStructureSet = SetCSPStructures([CSPStructure(i) for i in range(1,5)], calculation = 'crystalOptimiser')
newStructureSet.structures[2].setAttribute(Attribute(pointer={'crystalOptimiserData': 'rhoFinal'},
value = 23.4))
self.assertEqual(newStructureSet.structures[2].getAttribute({'crystalOptimiserData': 'rhoFinal'}),
23.4)
newStructureSet.setGlobalAttributes([Attribute(pointer={'crystalOptimiserData': 'rhoFinal'},
value=1.0)])
self.assertEqual(newStructureSet.structures[2].getAttribute({'crystalOptimiserData': 'rhoFinal'}),
1.)
self.assertEqual(newStructureSet.calculation,
'crystalOptimiser')
# print newStructureSet.structures[0].__dict__
print newStructureSet.statusString()
#read some PCM data
newStructureSet.addInfoFromGlobString('1-1/PCM/1-1', calculation = 'pcm')
self.assertAlmostEqual(newStructureSet.structures[0].pcmData['eInterFinal'],
-149.0522)
self.assertAlmostEqual(newStructureSet.structures[0].pcmData['eIntraHartrees'],
-1043.20561970)
self.assertTrue(newStructureSet.structures[0].pcmData['finalStructureFileExists'])
def tearDown(self):
print 'tearing down', os.path.isfile('tempCIFWrite.cif')
def rmsdtime(k, m):
rpt = 0
xax = []
process_time_list = []
compare_time_list = []
total_time_list = []
while rpt <= m:
file_extract = file_select(k)
filename_split = [i.split("/") for i in file_extract]
filename_list = [str(i[len(i) - 1][4:-4]) for i in filename_split]
print("Files to be compared:")
print(filename_list)
a = time.perf_counter()
crystals = [CrystalReader(c)[0] for c in file_extract]
#packed_crystals = [c.packing() for c in crystals]
b = time.perf_counter()
process_time_list.append(b - a)
print(f'Processing time (s) for %s molecules:', k)
print((b - a))
print(process_time_list)
c = time.perf_counter()
for i, j in itertools.combinations_with_replacement(
range(len(crystals)), 2):
comp = similarity_engine.compare(crystals[i], crystals[j])
d = time.perf_counter()
print(f'Comparison time (s) for %s molecules:', k)
print((d - c) + (a - b))
compare_time_list.append(d - c)
print([compare_time_list])
total_time_list.append((b - a) + (d - c))
print([total_time_list])
rpt += 1
xax.append(rpt)
plt.plot(xax, process_time_list, label='Processing Time')
plt.plot(xax, compare_time_list, label='Comparison Time')
plt.plot(xax, compare_time_list, label='Total Time')
plt.xlabel('Run number')
plt.ylabel('Computation time')
plt.title(f'RMSD Comparison')
plt.legend()
plt.savefig(outputdir + "/RMSD_time_test.png")
plt.cla()
print(f'Average rmsd processing time (s) for 5 molecules:')
print(sum(process_time_list) / len(process_time_list))
print(f'Average rmsed pairwise comparison time for 5 molecules')
print(sum(compare_time_list) / len(compare_time_list))
print(f'Total rmsd time for %s molecules', k)
print(sum(total_time_list) / len(total_time_list))
return [
sum(process_time_list) / len(process_time_list),
sum(compare_time_list) / len(compare_time_list),
sum(total_time_list) / len(total_time_list)
]
def powdertime(k, m):
rpt = 0
xax = []
process_time_list = []
compare_time_list = []
total_time_list = []
while rpt <= m:
file_extract = file_select(k)
#filename_split = [i.split("/") for i in file_extract]
#filename_list = [str(i[len(i)-1][4:-4]) for i in filename_split]
a = time.perf_counter()
crystals = [CrystalReader(c)[0] for c in file_extract]
#packed_crystals = [c.packing() for c in crystals]
powders = [CD.PowderPattern.from_crystal(c) for c in crystals]
b = time.perf_counter()
process_time_list.append(b - a)
print(f'Processing time (s) for 5 molecules')
print((b - a))
c = time.perf_counter()
for i, j in itertools.combinations_with_replacement(
range(len(crystals)), 2):
powd = powders[i].similarity(powders[j])
d = time.perf_counter()
compare_time_list.append(d - c)
total_time_list.append((b - a) + (d - c))
print(f'Compariosn time (s) for 5 molecules:')
rpt += 1
xax.append(rpt)
plt.plot(xax, process_time_list, label='Processing Time')
plt.plot(xax, compare_time_list, label='Comparison Time')
plt.plot(xax, total_time_list, label='Total Time')
plt.xlabel('Run number')
plt.ylabel('Computation time')
plt.title('Powder Comparison')
plt.legend()
plt.savefig(outputdir + "/powder_time_test.png")
plt.cla()
print(f'Average powder processing time (s) for %s molecules:', k)
print(sum(process_time_list) / len(process_time_list))
print(f'Average powder pairwise comparison time for %s molecules', k)
print(sum(compare_time_list) / len(compare_time_list))
print(f'Total powder time for %s molecules', k)
print(sum(total_time_list) / len(total_time_list))
return [
sum(process_time_list) / len(process_time_list),
sum(compare_time_list) / len(compare_time_list),
sum(total_time_list) / len(total_time_list)
]
try:
return compareResult.nmatched_molecules
except:
return None
# Set up comparison object
ps = PackingSimilarity()
allowMolecularDifferences = True
clusterSize = 15
ps.settings.allow_molecular_differences = allowMolecularDifferences
ps.settings.packing_shell_size = clusterSize
# define crystals
try:
crystal1 = CrystalReader(sys.argv[1], format='res')[0]
except:
crystal1 = CrystalReader(sys.argv[1], format='cif')[0]
try:
crystal2 = CrystalReader(sys.argv[2], format='res')[0]
except:
crystal2 = CrystalReader(sys.argv[2], format='cif')[0]
matchingData = returnNmatched_molecules(ps,
crystal1,
crystal2,
returnRMSD=True)
#print "Matching {1} molecules (out of {2}) with RMSD = {3} Angstroms".format(clusterSize,clusterSize,clusterSize)
print "Allowing molecular differences? ", allowMolecularDifferences
print "Matching %s molecules (out of %s) with RMSD = %s Angstroms" % (
link_list = []
print(st + e)
for file_path in files[st:e]:
file_path_split = file_path.split("/")
file_name = file_path_split[len(file_path_split) - 1]
file_list.append(file_name)
doc = cif.read(file_path)
block = doc[0]
for b in doc:
if (b.find_loop('_atom_site_') != None):
block = b
print("Processed file --> " + file_name)
crystal_reader = CrystalReader(file_path)
crystal = crystal_reader[0]
crystal.assign_bonds()
packed_molecules = crystal.packing(box_dimensions=((0, 0, 0), (1, 1, 1)),
inclusion='CentroidIncluded')
packed_molecules.normalise_labels()
adta_molecules = []
cent = []
cent_points = []
for comp in packed_molecules.components:
if (len(comp.atoms) > 1):
adta_molecules.append(comp)
cent.append(MD.atom_centroid(*list(a for a in comp.atoms)))
for c in cent:
from ase.build import niggli_reduce
from ase import Atoms
from ase.io import write as aseWrite
from ioAndInterfaces import ccdcCrystalToASE
from ccdc.io import CrystalReader
#example input and outputs to test this
inputRes, outputRes = 'testingScripts/new.res', 'temp070917.res'
myCell = CrystalReader(inputRes)[0]
myASECell = ccdcCrystalToASE(myCell)
myASECell.set_pbc(True)
niggli_reduce(myASECell)
aseWrite(outputRes, myASECell)
print 'issue with this is that crystal optimiser wants labels etc, and probably fussy with res file format'
#could attach labels or something if can be bothered
print 'another issue is that the cell may not have the same convention, so would have to reset rather
than just changing angles etc'
for pair in MolecularDescriptors.MaximumCommonSubstructure().search(mol1, mol2)[0]])
def main():
try:
# oldString, oldResFile, templateResFile = sys.argv[1:]
oldString, resFile = sys.argv[1:]
except:
raise Exception(usefulMessage)
# print simpleRelabelling("+O1_C1_C2_C7 +H1_O1_C1_C2 +C1_C2_C7_N1 +C7_N1_C8_C9 +H6_N1_C8_C9 +F1_C14_C12_C11 +H6_N1_C8 +H1_O1_C1",
print simpleRelabelling("O1_C1_C2_C7 H1_O1_C1_C2 C1_C2_C7_N1 C7_N1_C8_C9 H6_N1_C8_C9 F1_C14_C12_C11 H6_N1_C8 H1_O1_C1",
CrystalReader(resFile)[0])
if __name__ == '__main__':
main()
exit()
print relabelString("+O1_C1_C2_C7 +H1_O1_C1_C2 +C1_C2_C7_N1 +C7_N1_C8_C9 +H6_N1_C8_C9 +F1_C14_C12_C11 +H6_N1_C8 +H1_O1_C1",
replacementDict(CrystalReader(templateResFile)[0],
CrystalReader(oldResFile)[0]))
exit()
exit()
# use the above subroutines to achieve the relabelling
# uses ccdc molecule instances
with open(newFilename, 'w') as outf:
outf.write(reorderedRes(oldResFile, replacementDict(CrystalReader(templateResFile)[0],
CrystalReader(oldResFile)[0]))
)
def linkcomp(start, end, pcomp=True, rcomp=True, lktest=False):
full_list = []
file_extract = files[start:end]
filename_split = [i.split("/") for i in file_extract]
filename_list = [str(i[len(i) - 1][4:-4]) for i in filename_split]
if lktest:
lkarray = np.zeros((len(file_extract), len(file_extract)))
if pcomp:
powdarray = np.zeros((len(file_extract), len(file_extract)))
if rcomp:
rmsdarray = np.zeros((len(file_extract), len(file_extract)))
for first_comp in file_extract:
i = file_extract.index(first_comp)
rest_list = file_extract[i + 1:end]
new_entry = []
for second_comp in rest_list:
j = file_extract.index(second_comp)
first_comp_split = first_comp.split("/")
second_comp_split = second_comp.split("/")
one_crys_read = CrystalReader(first_comp)
two_crys_read = CrystalReader(second_comp)
one_crys = one_crys_read[0]
two_crys = two_crys_read[0]
powder_sim = CD.PowderPattern.from_crystal(one_crys)
powder_comp = CD.PowderPattern.from_crystal(two_crys)
comp = similarity_engine.compare(one_crys, two_crys)
powd = powder_sim.similarity(powder_comp)
if lktest:
try:
lkarray[j][i] += (round(
abs(
max(link_dic[first_file_name]) -
max(link_dic[second_file_name])), 3))
except TypeError:
lkarray[j][i] += np.nan
if pcomp:
if powd is None:
powdarray[j][i] += np.nan
else:
powdarray[j][i] += (round(powd, 3))
if rcomp:
if comp is None:
rmsdarray[j][i] += np.nan
else:
rmsdarray[j][i] += (round(comp.rmsd, 3))
if lktest:
lkframe = pd.DataFrame(lkarray,
index=filename_list,
columns=filename_list)
lkframe.to_csv(outputdir + "/link_comparison.csv",
index=True,
header=True,
sep=',')
if pcomp:
powdframe = pd.DataFrame(powdarray,
index=filename_list,
columns=filename_list)
powdframe.to_csv(outputdir + "/powder_comparison.csv",
index=True,
header=True,
sep=',')
if rcomp:
powdframe = pd.DataFrame(rmsdarray,
index=filename_list,
columns=filename_list)
powdframe.to_csv(outputdir + "/rmsd_comparison.csv",
index=True,
header=True,
sep=',')
def labelIncludingNumberMols(atom1Label, stoichiometry, molIndex):
element1, number1 = re.split('(\d+)', atom1Label)[:2]
return element1 + str(
int(number1) + int(stoichiometry[element1]) * molIndex)
blurb = '''This file contains the Z-matrix definition for each of the molecules
in the asymmetric unit of the crystal to be minimised by CrystalOptimizer.
It is absolutely necessary to contain an 'introductory' line for the Z-matrix
of each molecule followed imidiately (no empty line) by the Z-matrix
definition. The introductory line should be 'Z-matrix for molecule'
followed by the enumeration of teh molecule. The Z-matrix definition
should be done using the atom labels in exactly the same way as they
appear in the .res file.'''
crystal = CrystalReader(sys.argv[1])[0]
stoichiometry = dict([
re.split('(\d+)', x)[:2]
for x in crystal.molecule.components[0].formula.split()
])
zMatrixString = open(sys.argv[2], 'r').read()
print blurb #zMatrixString
for i in xrange(len(crystal.molecule.components)):
print "\nZ-matrix for molecule %s" % (i + 1)
for l in zMatrixString.split('\n')[1:]:
# print l
print " ".join(
[labelIncludingNumberMols(x, stoichiometry, i) for x in l.split()])
def molcomp(start,
end,
pcomp=True,
rmsd_threshold=2,
rounding=6,
rcomp=True,
molcomp=True):
filename_split = [i.split("/") for i in file_extract]
filename_list = [str(i[len(i) - 1][4:-4]) for i in filename_split]
l = len(file_extract)
if molcomp:
molarray = np.zeros((l, l))
if pcomp:
powdarray = np.zeros((l, l))
if rcomp:
rmsdarray = np.zeros((l, l))
crystals = [CrystalReader(c)[0] for c in file_extract]
powders = [CD.PowderPattern.from_crystal(c) for c in crystals]
for i, j in itertools.combinations_with_replacement(
range(len(crystals)), 2):
comp = similarity_engine.compare(crystals[i], crystals[j])
powd = powders[i].similarity(powders[j])
if molcomp:
try:
molarray[j][i] += (comp.nmatched_molecules)
molarray[i][j] += (comp.nmatched_molecules)
except TypeError:
molarray[j][i] += 99
molarray[i][j] += 99
if pcomp:
if powd is None:
powdarray[j][i] += 99
powfarray[i][j] += 99
else:
powdarray[j][i] += (round(powd, 6))
powdarray[i][j] += (round(powd, 6))
if rcomp:
if comp is None:
rmsdarray[j][i] += 99
rmsdarray[i][j] += 99
elif comp.nmatched_molecules < rmsd_threshold:
rmsdarray[j][i] += 88
rmsdarray[i][j] += 88
else:
rmsdarray[j][i] += (round(comp.rmsd, 6))
rmsdarray[i][j] += (round(comp.rmsd, 6))
if molcomp:
molframe = pd.DataFrame(molarray,
index=filename_list,
columns=filename_list)
molframe.replace(99, 'nan')
molframe.to_csv(outputdir + "/rmsd_molecules_matched.csv",
index=True,
header=True,
sep=',')
if pcomp:
powdframe = pd.DataFrame(powdarray,
index=filename_list,
columns=filename_list)
powdframe.to_csv(outputdir + "/powder_comparison.csv",
index=True,
header=True,
sep=',')
if rcomp:
rmsdframe = pd.DataFrame(rmsdarray,
index=filename_list,
columns=filename_list)
rmsdframe.to_csv(outputdir + "/rmsd_comparison.csv",
index=True,
header=True,
sep=',')