def find_unconnected_xds_files(xds_files, min_ios=3, d_min=None):
import networkx as nx
from yamtbx.dataproc.xds import xds_ascii
from yamtbx.dataproc.xds import integrate_hkl_as_flex
if len(xds_files) < 2:
return []
G = nx.Graph()
arrays = []
for f in xds_files:
if xds_ascii.is_xds_ascii(f):
a = xds_ascii.XDS_ASCII(f, i_only=True).i_obs().resolution_filter(d_min=d_min)
elif integrate_hkl_as_flex.is_integrate_hkl(f):
a = integrate_hkl_as_flex.reader(f, ["IOBS", "SIGMA"]).i_obs().resolution_filter(d_min=d_min)
else:
raise "Never reaches here"
a = a.select(a.sigmas() > 0)
a = a.merge_equivalents(use_internal_variance=False).array()
arrays.append(a.select(a.data() / a.sigmas() >= min_ios))
for i in xrange(len(arrays) - 1):
for j in xrange(i + 1, len(arrays)):
matchs = arrays[i].match_indices(other=arrays[j], assert_is_similar_symmetry=False)
if matchs.pairs().size() >= 10:
G.add_edge(i, j)
print "edge", i, j
ccomps = map(lambda x: x, nx.connected_components(G))
print "DEBUG:: Connected components=", ccomps
keep_idxes = ccomps[0]
remove_idxes = filter(lambda x: x not in keep_idxes, xrange(len(xds_files)))
return remove_idxes
def load_xds_data_only_indices(xac_files, d_min=None):
miller_sets = {}
for f in xac_files:
if xds_ascii.is_xds_ascii(f):
print "Loading", f
ms = xds_ascii.XDS_ASCII(f, i_only=True).as_miller_set()
miller_sets[f] = ms.resolution_filter(d_min=d_min)
elif integrate_hkl_as_flex.is_integrate_hkl(f):
print "Sorry, skipping", f
else:
print "Skipping unrecognized:", f
return miller_sets
def load_xds_data_only_indices(xac_files, d_min=None):
miller_sets = {}
for f in xac_files:
if xds_ascii.is_xds_ascii(f):
print "Loading", f
ms = xds_ascii.XDS_ASCII(f, i_only=True).as_miller_set()
miller_sets[f] = ms.resolution_filter(d_min=d_min)
elif integrate_hkl_as_flex.is_integrate_hkl(f):
print "Sorry, skipping", f
else:
print "Skipping unrecognized:", f
return miller_sets
def load_xds_data_only_indices(xac_files, d_min=None):
miller_sets = {}
for f in xac_files:
try:
indices = []
miller_arrays = any_reflection_file(f).as_miller_arrays()
symm = miller_arrays[0].crystal_symmetry()
ms = miller.set(crystal_symmetry=symm, indices=miller_arrays[0].indices(),anomalous_flag=False)
miller_sets[f] = ms.resolution_filter(d_min=d_min)
except:
if xds_ascii.is_xds_ascii(f):
print "Loading", f
ms = xds_ascii.XDS_ASCII(f, i_only=True).as_miller_set()
miller_sets[f] = ms.resolution_filter(d_min=d_min)
elif integrate_hkl_as_flex.is_integrate_hkl(f):
print "Sorry, skipping", f
else:
print "Skipping unrecognized:", f
return miller_sets
def find_unconnected_xds_files(xds_files, min_ios=3, d_min=None):
import networkx as nx
from yamtbx.dataproc.xds import xds_ascii
from yamtbx.dataproc.xds import integrate_hkl_as_flex
if len(xds_files) < 2:
return []
G = nx.Graph()
arrays = []
for f in xds_files:
if xds_ascii.is_xds_ascii(f):
a = xds_ascii.XDS_ASCII(
f, i_only=True).i_obs().resolution_filter(d_min=d_min)
elif integrate_hkl_as_flex.is_integrate_hkl(f):
a = integrate_hkl_as_flex.reader(
f, ["IOBS", "SIGMA"]).i_obs().resolution_filter(d_min=d_min)
else:
raise "Never reaches here"
a = a.select(a.sigmas() > 0)
a = a.merge_equivalents(use_internal_variance=False).array()
arrays.append(a.select(a.data() / a.sigmas() >= min_ios))
for i in xrange(len(arrays) - 1):
for j in xrange(i + 1, len(arrays)):
matchs = arrays[i].match_indices(other=arrays[j],
assert_is_similar_symmetry=False)
if matchs.pairs().size() >= 10:
G.add_edge(i, j)
print "edge", i, j
ccomps = map(lambda x: x, nx.connected_components(G))
print "DEBUG:: Connected components=", ccomps
keep_idxes = ccomps[0]
remove_idxes = filter(lambda x: x not in keep_idxes,
xrange(len(xds_files)))
return remove_idxes
an, bn = params.variance_model
i_obs = i_obs.customized_copy(sigmas = flex.sqrt(flex.abs(an * (i_obs.sigmas()**2/ao + (bn-bo)*flex.pow2(i_obs.data())))))
else:
ihkl = integrate_hkl_as_flex.reader(f, read_columns=("IOBS","SIGMA"))
i_obs = ihkl.i_obs().resolution_filter(d_min=params.d_min, d_max=params.d_max)
if params.fix_variance_model:
a, b = params.variance_model
i_obs = i_obs.customized_copy(sigmas = flex.sqrt(a * (i_obs.sigmas()**2 + b*flex.pow2(i_obs.data()))))
cutoffs = eval_resolution(i_obs, params.n_bins, params.cut_ios)
print "%s %s" % (f, " ".join(map(lambda x: "%.2f"%x, cutoffs)))
if __name__ == "__main__":
import sys
cmdline = iotbx.phil.process_command_line(args=sys.argv[1:],
master_string=master_params_str)
params = cmdline.work.extract()
args = cmdline.remaining_args
files = filter(lambda x: integrate_hkl_as_flex.is_integrate_hkl(x) or xds_ascii.is_xds_ascii(x), args)
for arg in args:
if arg.endswith(".lst"):
for l in open(arg):
if "#" in l: l = l[:l.index("#")]
files.append(l.strip())
run(files, params)
