def _prepare_pointless_hklin(working_directory,
hklin,
phi_width):
'''Prepare some data for pointless - this will take only 180 degrees
of data if there is more than this (through a "rebatch" command) else
will simply return hklin.'''
# also remove blank images?
if not Flags.get_microcrystal() and not Flags.get_small_molecule():
Debug.write('Excluding blank images')
hklout = os.path.join(
working_directory,
'%s_noblank.mtz' % (os.path.split(hklin)[-1][:-4]))
FileHandler.record_temporary_file(hklout)
hklin = remove_blank(hklin, hklout)
# find the number of batches
md = Mtzdump()
md.set_working_directory(working_directory)
auto_logfiler(md)
md.set_hklin(hklin)
md.dump()
batches = max(md.get_batches()) - min(md.get_batches())
phi_limit = 180
if batches * phi_width < phi_limit or Flags.get_small_molecule():
return hklin
hklout = os.path.join(
working_directory,
'%s_prepointless.mtz' % (os.path.split(hklin)[-1][:-4]))
rb = Rebatch()
rb.set_working_directory(working_directory)
auto_logfiler(rb)
rb.set_hklin(hklin)
rb.set_hklout(hklout)
first = min(md.get_batches())
last = first + int(phi_limit / phi_width)
Debug.write('Preparing data for pointless - %d batches (%d degrees)' % \
((last - first), phi_limit))
rb.limit_batches(first, last)
# we will want to delete this one exit
FileHandler.record_temporary_file(hklout)
return hklout
def copyfree(self):
'''Copy the free column from freein into hklin -> hklout.'''
if not self._hklin_files:
raise RuntimeError('no hklin files defined')
if len(self._hklin_files) > 1:
raise RuntimeError('can have only one hklin to update')
hklin = self._hklin_files[0]
# get the resolution limit to give as a limit for the FreeR
# column
md = Mtzdump()
md.set_working_directory(self.get_working_directory())
md.set_hklin(hklin)
md.dump()
resolution_range = md.get_resolution_range()
self.check_hklout()
if self._freein is None:
raise RuntimeError('freein not defined')
if self._freein_column is None:
raise RuntimeError('freein column not defined')
self.add_command_line('hklin1')
self.add_command_line(self._freein)
self.add_command_line('hklin2')
self.add_command_line(hklin)
self.start()
self.input('labin file_number 1 E1=%s' % self._freein_column)
self.input('resolution file_number 1 %f %f' % resolution_range)
self.input('labin file_number 2 all')
self.close_wait()
try:
self.check_for_errors()
self.check_ccp4_errors()
except RuntimeError as e:
# something went wrong; remove the output file
try:
os.remove(self.get_hklout())
except Exception:
pass
raise e
return self.get_ccp4_status()
def FindFreeFlag(hklin):
"""Try to find the FREE column in hklin. Raise exception if no column is
found or if more than one candidate is found."""
# get the information we need here...
mtzdump = Mtzdump()
mtzdump.set_hklin(hklin)
mtzdump.dump()
columns = mtzdump.get_columns()
ctypes = {c[0]: c[1] for c in columns}
if "FreeR_flag" in ctypes:
if ctypes["FreeR_flag"] != "I":
raise RuntimeError("FreeR_flag column found: type not I")
return "FreeR_flag"
# ok, so the usual one wasn't there, look for anything with "free"
# in it...
possibilities = [c for c in ctypes if "free" in c.lower()]
if not possibilities:
raise RuntimeError("no candidate FreeR_flag columns found")
if len(possibilities) == 1:
if ctypes[possibilities[0]] != "I":
raise RuntimeError("FreeR_flag column found (%s): type not I" %
possibilities[0])
return possibilities[0]
def find_columns(self):
'''Identify columns to use with scaleit.'''
# run mtzdump to get a list of columns out and also check that
# this is a valid merged mtz file....
self.check_hklin()
md = Mtzdump()
md.set_hklin(self.get_hklin())
md.dump()
# get information to check that this is merged
# next get the column information - check that F columns are
# present
column_info = md.get_columns()
columns = []
j = 0
groups = 0
# assert that the columns for F, SIGF, DANO, SIGDANO for a
# particular group will appear in that order if anomalous,
# F, SIGF if not anomalous
while j < len(column_info):
c = column_info[j]
name = c[0]
type = c[1]
if type == 'F' and name.split('_')[0] == 'F' and \
self._anomalous:
groups += 1
for i in range(4):
columns.append(column_info[i + j][0])
j += 4
elif type == 'F' and name.split('_')[0] == 'F' and \
not self._anomalous:
groups += 1
for i in range(2):
columns.append(column_info[i + j][0])
j += 2
else:
j += 1
# ok that should be all of the groups identified
self._columns = columns
return columns
def FindFreeFlag(hklin):
'''Try to find the FREE column in hklin. Raise exception if no column is
found or if more than one candidate is found.'''
# get the information we need here...
mtzdump = Mtzdump()
mtzdump.set_hklin(hklin)
mtzdump.dump()
columns = mtzdump.get_columns()
ctypes = {}
for c in columns:
ctypes[c[0]] = c[1]
if 'FreeR_flag' in ctypes.keys():
if ctypes['FreeR_flag'] != 'I':
raise RuntimeError('FreeR_flag column found: type not I')
return 'FreeR_flag'
# ok, so the usual one wasn't there, look for anything with "free"
# in it...
possibilities = []
for c in ctypes.keys():
if 'free' in c.lower():
possibilities.append(c)
if len(possibilities) == 0:
raise RuntimeError('no candidate FreeR_flag columns found')
if len(possibilities) == 1:
if ctypes[possibilities[0]] != 'I':
raise RuntimeError('FreeR_flag column found (%s): type not I' % \
possibilities[0])
return possibilities[0]
raise RuntimeError('Multiple candidate FreeR_flag columns found')
def FindFreeFlag(hklin):
'''Try to find the FREE column in hklin. Raise exception if no column is
found or if more than one candidate is found.'''
# get the information we need here...
mtzdump = Mtzdump()
mtzdump.set_hklin(hklin)
mtzdump.dump()
columns = mtzdump.get_columns()
ctypes = { }
for c in columns:
ctypes[c[0]] = c[1]
if 'FreeR_flag' in ctypes.keys():
if ctypes['FreeR_flag'] != 'I':
raise RuntimeError, 'FreeR_flag column found: type not I'
return 'FreeR_flag'
# ok, so the usual one wasn't there, look for anything with "free"
# in it...
possibilities = []
for c in ctypes.keys():
if 'free' in c.lower():
possibilities.append(c)
if len(possibilities) == 0:
raise RuntimeError, 'no candidate FreeR_flag columns found'
if len(possibilities) == 1:
if ctypes[possibilities[0]] != 'I':
raise RuntimeError, 'FreeR_flag column found (%s): type not I' % \
possibilities[0]
return possibilities[0]
raise RuntimeError, 'Multiple candidate FreeR_flag columns found'
def update(self):
'''Update the information for one reflection file.'''
if not self._hklin_files:
raise RuntimeError('no hklin files defined')
if len(self._hklin_files) > 1:
raise RuntimeError('can have only one hklin to update')
hklin = self._hklin_files[0]
self.check_hklout()
column_names_by_file = {}
dataset_names_by_file = {}
md = Mtzdump()
md.set_hklin(hklin)
md.dump()
columns = md.get_columns()
column_names_by_file[hklin] = []
dataset_names_by_file[hklin] = md.get_datasets()
# get a dataset ID - see FIXME 03/NOV/06 below...
dataset_ids = [md.get_dataset_info(d)['id'] for \
d in md.get_datasets()]
for c in columns:
name = c[0]
if name in ['H', 'K', 'L']:
continue
column_names_by_file[hklin].append(name)
self.add_command_line('hklin1')
self.add_command_line(hklin)
self.start()
dataset_id = dataset_ids[0]
if self._pname and self._xname and self._dname:
self.input('drename file_number 1 %d %s %s' % \
(dataset_id, self._xname, self._dname))
self.input('dpname file_number 1 %d %s' % \
(dataset_id, self._pname))
column_counter = 0
labin_command = 'labin file_number 1'
for column in column_names_by_file[hklin]:
column_counter += 1
labin_command += ' E%d=%s' % (column_counter, column)
self.input(labin_command)
# FIXME perhaps - ASSERT that we want only the information from
# the first dataset here...
pname, xname, dname = dataset_names_by_file[hklin][0].split('/')
dataset_id = dataset_ids[0]
# FIXME 03/NOV/06 this needs to id the dataset by it's number
# not by pname/xname/dname, as the latter get's confused if the
# xname is a number...
if self._new_cell_parameters:
a, b, c, alpha, beta, gamma = self._new_cell_parameters
self.input('dcell file_number 1 %d %f %f %f %f %f %f' % \
(dataset_id, a, b, c, alpha, beta, gamma))
if self._new_column_suffix:
suffix = self._new_column_suffix
column_counter = 0
labout_command = 'labout file_number 1'
for column in column_names_by_file[hklin]:
column_counter += 1
labout_command += ' E%d=%s_%s' % \
(column_counter, column, suffix)
self.input(labout_command)
self.close_wait()
try:
self.check_for_errors()
self.check_ccp4_errors()
except RuntimeError as e:
# something went wrong; remove the output file
try:
os.remove(self.get_hklout())
except Exception:
pass
raise e
return self.get_ccp4_status()
def merge(self):
'''Merge multiple reflection files into one file.'''
if not self._hklin_files:
raise RuntimeError('no hklin files defined')
self.check_hklout()
hklin_counter = 0
# for each reflection file, need to gather the column names
# and so on, to put in the cad input here - also check to see
# if the column names clash... check also that the spacegroups
# match up...
spacegroup = None
column_names = []
column_names_by_file = {}
for hklin in self._hklin_files:
md = Mtzdump()
md.set_working_directory(self.get_working_directory())
md.set_hklin(hklin)
md.dump()
columns = md.get_columns()
spag = md.get_spacegroup()
if spacegroup is None:
spacegroup = spag
if spag != spacegroup:
raise RuntimeError('spacegroups do not match')
column_names_by_file[hklin] = []
for c in columns:
name = c[0]
if name in ['H', 'K', 'L']:
continue
if name in column_names:
raise RuntimeError('duplicate column names')
column_names.append(name)
column_names_by_file[hklin].append(name)
# if we get to here then this is a good set up...
# create the command line
hklin_counter = 0
for hklin in self._hklin_files:
hklin_counter += 1
self.add_command_line('hklin%d' % hklin_counter)
self.add_command_line(hklin)
self.start()
hklin_counter = 0
for hklin in self._hklin_files:
column_counter = 0
hklin_counter += 1
labin_command = 'labin file_number %d' % hklin_counter
for column in column_names_by_file[hklin]:
column_counter += 1
labin_command += ' E%d=%s' % (column_counter, column)
self.input(labin_command)
self.close_wait()
try:
self.check_for_errors()
self.check_ccp4_errors()
except RuntimeError as e:
# something went wrong; remove the output file
try:
os.remove(self.get_hklout())
except Exception:
pass
raise e
return self.get_ccp4_status()
if __name__ == '__main__':
# ok, in here (which will be "autoCHEF") this will inspect the MTZ
# file and run with DOSE if such a column exists, else will run with
# BATCH. N.B. this will require a fix above.
chef = Chef(stream = Stdout)
dose_column = None
overall_dmin = None
for argv in sys.argv[1:]:
md = Mtzdump()
md.set_hklin(argv)
md.dump()
columns = [c[0] for c in md.get_columns()]
if dose_column:
assert(dose_column in columns)
continue
if 'DOSE' in columns:
dose_range = md.get_column_range('DOSE')[:2]
if dose_range[0] != dose_range[1]:
dose_column = 'DOSE'
if __name__ == '__main__':
# ok, in here (which will be "autoCHEF") this will inspect the MTZ
# file and run with DOSE if such a column exists, else will run with
# BATCH. N.B. this will require a fix above.
chef = Chef(stream = Stdout)
dose_column = None
overall_dmin = None
for argv in sys.argv[1:]:
md = Mtzdump()
md.set_hklin(argv)
md.dump()
columns = [c[0] for c in md.get_columns()]
if dose_column:
assert(dose_column in columns)
continue
if 'DOSE' in columns:
dose_range = md.get_column_range('DOSE')[:2]
if dose_range[0] != dose_range[1]:
dose_column = 'DOSE'
def update(self):
"""Update the information for one reflection file."""
if not self._hklin_files:
raise RuntimeError("no hklin files defined")
if len(self._hklin_files) > 1:
raise RuntimeError("can have only one hklin to update")
hklin = self._hklin_files[0]
self.check_hklout()
column_names_by_file = {}
dataset_names_by_file = {}
md = Mtzdump()
md.set_hklin(hklin)
md.dump()
columns = md.get_columns()
column_names_by_file[hklin] = []
dataset_names_by_file[hklin] = md.get_datasets()
# get a dataset ID - see FIXME 03/NOV/06 below...
dataset_ids = [md.get_dataset_info(d)["id"] for d in md.get_datasets()]
for c in columns:
name = c[0]
if name in ["H", "K", "L"]:
continue
column_names_by_file[hklin].append(name)
self.add_command_line("hklin1")
self.add_command_line(hklin)
self.start()
dataset_id = dataset_ids[0]
if self._pname and self._xname and self._dname:
self.input(
"drename file_number 1 %d %s %s"
% (dataset_id, self._xname, self._dname)
)
self.input("dpname file_number 1 %d %s" % (dataset_id, self._pname))
column_counter = 0
labin_command = "labin file_number 1"
for column in column_names_by_file[hklin]:
column_counter += 1
labin_command += " E%d=%s" % (column_counter, column)
self.input(labin_command)
# FIXME perhaps - ASSERT that we want only the information from
# the first dataset here...
dataset_id = dataset_ids[0]
if self._new_cell_parameters:
a, b, c, alpha, beta, gamma = self._new_cell_parameters
self.input(
"dcell file_number 1 %d %f %f %f %f %f %f"
% (dataset_id, a, b, c, alpha, beta, gamma)
)
if self._new_column_suffix:
suffix = self._new_column_suffix
column_counter = 0
labout_command = "labout file_number 1"
for column in column_names_by_file[hklin]:
column_counter += 1
labout_command += " E%d=%s_%s" % (column_counter, column, suffix)
self.input(labout_command)
self.close_wait()
try:
self.check_for_errors()
self.check_ccp4_errors()
except RuntimeError as e:
# something went wrong; remove the output file
try:
os.remove(self.get_hklout())
except Exception:
pass
raise e
return self.get_ccp4_status()