def main(args):
log = logging.getLogger('root')
hdlr = logging.StreamHandler(sys.stdout)
log.addHandler(hdlr)
log.setLevel(logging.getLevelName(args.loglevel.upper()))
dfs = [metadata.parse_fx_par(fn) for fn in args.input]
n = dfs[0].shape[0]
if not np.all(np.array([df.shape[0] for df in dfs]) == n):
log.error("Input files are not aligned!")
return 1
df = pd.concat(dfs, axis=0, ignore_index=True)
df["CLASS"] = np.repeat(np.arange(1, len(dfs) + 1), n)
if args.min_occ:
df = df[df["OCC"] >= args.min_occ]
df = df.sort_values(by="OCC")
df = df.drop_duplicates("C", keep="last")
df = df.sort_values(by="C")
df = metadata.par2star(df,
data_path=args.stack,
apix=args.apix,
cs=args.cs,
ac=args.ac,
kv=args.voltage,
invert_eulers=args.relion)
if args.cls is not None:
df = star.select_classes(df, args.cls)
star.write_star(args.output, df, reindex=True)
return 0
def main(args):
if args.data_path is None:
args.data_path = args.input.rstrip(".par") + ".mrcs"
print("No stack path provided, using %s" % args.data_path)
df = parse_fx_par(args.input)
df = par2star(df,
data_path=args.data_path,
apix=args.apix,
cs=args.cs,
ac=args.ac,
kv=args.voltage)
write_star(args.output, df, reindex=True)
return 0
def main(args):
par = [parse_par(inp) for inp in args.input]
if len(par) > 1:
idx = par[:, :, "OCC"].idxmax(axis=1)
else:
par = par[0]
if args.data_path is not None:
par["Input particle images"] = args.data_path
if args.voltage is not None:
par["Beam energy (keV)"] = args.voltage
if args.cs is not None:
par["Spherical aberration (mm)"] = args.cs
if args.ac is not None:
par["Amplitude contrast"] = args.ac
if args.apix is not None:
par["Pixel size of images (A)"] = args.apix
star = par2star(par)
write_star(args.output, star, reindex=True)
return 0
def main(args):
log = logging.getLogger('root')
hdlr = logging.StreamHandler(sys.stdout)
log.addHandler(hdlr)
log.setLevel(logging.getLevelName(args.loglevel.upper()))
dfs = [metadata.parse_fx_par(fn) for fn in args.input]
n = dfs[0].shape[0]
if not np.all(np.array([df.shape[0] for df in dfs]) == n):
log.error("Input files are not aligned!")
return 1
df = pd.concat(dfs, axis=0, ignore_index=True)
df["CLASS"] = np.repeat(np.arange(1, len(dfs) + 1), n)
if args.min_occ:
df = df[df["OCC"] >= args.min_occ]
df = df.sort_values(by="OCC")
df = df.drop_duplicates("C", keep="last")
df = df.sort_values(by="C")
df.reset_index(inplace=True)
if args.min_score is not None:
if args.min_score < 1:
args.min_score = np.percentile(df["SCORE"],
(1 - args.min_score) * 100)
df = df.loc[df["SCORE"] >= args.min_score]
if args.merge is not None:
dfo = star.parse_star(args.merge)
args.apix = star.calculate_apix(dfo)
args.cs = dfo.iloc[0][star.Relion.CS]
args.ac = dfo.iloc[0][star.Relion.AC]
args.voltage = dfo.iloc[0][star.Relion.VOLTAGE]
df = metadata.par2star(df,
data_path=args.stack,
apix=args.apix,
cs=args.cs,
ac=args.ac,
kv=args.voltage,
invert_eulers=args.invert_eulers)
if args.stack is None:
df[star.UCSF.IMAGE_INDEX] = dfo[star.UCSF.IMAGE_INDEX]
df[star.UCSF.IMAGE_PATH] = dfo[star.UCSF.IMAGE_PATH]
key = [star.UCSF.IMAGE_INDEX, star.UCSF.IMAGE_PATH]
fields = star.Relion.MICROGRAPH_COORDS + [
star.UCSF.IMAGE_ORIGINAL_INDEX, star.UCSF.IMAGE_ORIGINAL_PATH
] + [star.Relion.OPTICSGROUP
] + star.Relion.OPTICSGROUPTABLE + [star.Relion.RANDOMSUBSET]
df = star.smart_merge(df, dfo, fields=fields, key=key)
if args.revert_original:
df = star.revert_original(df, inplace=True)
else:
df = metadata.par2star(df,
data_path=args.stack,
apix=args.apix,
cs=args.cs,
ac=args.ac,
kv=args.voltage,
invert_eulers=args.invert_eulers)
if args.cls is not None:
df = star.select_classes(df, args.cls)
df = star.check_defaults(df, inplace=True)
df = star.compatible(df, relion2=args.relion2, inplace=True)
star.write_star(args.output, df, optics=(not args.relion2))
return 0
def main(args):
log = logging.getLogger('root')
hdlr = logging.StreamHandler(sys.stdout)
log.addHandler(hdlr)
log.setLevel(logging.getLevelName(args.loglevel.upper()))
# apix = args.apix = hdr["xlen"] / hdr["nx"]
for fn in args.input:
if not (fn.endswith(".star") or fn.endswith(".mrcs")
or fn.endswith(".mrc") or fn.endswith(".par")):
log.error("Only .star, .mrc, .mrcs, and .par files supported")
return 1
first_ptcl = 0
dfs = []
with mrc.ZSliceWriter(args.output) as writer:
for fn in args.input:
if fn.endswith(".star"):
df = star.parse_star(fn, keep_index=False)
star.augment_star_ucsf(df)
star.set_original_fields(df, inplace=True)
df = df.sort_values([
star.UCSF.IMAGE_ORIGINAL_PATH,
star.UCSF.IMAGE_ORIGINAL_INDEX
])
gb = df.groupby(star.UCSF.IMAGE_ORIGINAL_PATH)
for name, g in gb:
with mrc.ZSliceReader(name) as reader:
for i in g[star.UCSF.IMAGE_ORIGINAL_INDEX].values:
writer.write(reader.read(i))
elif fn.endswith(".par"):
if args.stack_path is None:
log.error(".par file input requires --stack-path")
return 1
df = metadata.par2star(metadata.parse_fx_par(fn),
data_path=args.stack_path)
# star.set_original_fields(df, inplace=True) # Redundant.
star.augment_star_ucsf(df)
elif fn.endswith(".csv"):
return 1
elif fn.endswith(".cs"):
return 1
else:
if fn.endswith(".mrcs"):
with mrc.ZSliceReader(fn) as reader:
for img in reader:
writer.write(img)
df = pd.DataFrame({
star.UCSF.IMAGE_ORIGINAL_INDEX:
np.arange(reader.nz)
})
df[star.UCSF.IMAGE_ORIGINAL_PATH] = fn
else:
print("Unrecognized input file type")
return 1
if args.star is not None:
df[star.UCSF.IMAGE_INDEX] = np.arange(first_ptcl,
first_ptcl + df.shape[0])
df[star.UCSF.IMAGE_PATH] = writer.path
df["index"] = df[star.UCSF.IMAGE_INDEX]
star.simplify_star_ucsf(df)
dfs.append(df)
first_ptcl += df.shape[0]
if args.star is not None:
df = pd.concat(dfs, join="inner")
# df = pd.concat(dfs)
# df = df.dropna(df, axis=1, how="any")
star.write_star(args.star, df, reindex=True)
return 0
def main(args):
log = logging.getLogger('root')
hdlr = logging.StreamHandler(sys.stdout)
log.addHandler(hdlr)
log.setLevel(logging.getLevelName(args.loglevel.upper()))
# apix = args.apix = hdr["xlen"] / hdr["nx"]
for fn in args.input:
if not (fn.endswith(".star") or fn.endswith(".mrcs") or
fn.endswith(".mrc") or fn.endswith(".par")):
log.error("Only .star, .mrc, .mrcs, and .par files supported")
return 1
first_ptcl = 0
dfs = []
with mrc.ZSliceWriter(args.output) as writer:
for fn in args.input:
if fn.endswith(".star"):
df = star.parse_star(fn, augment=True)
if args.cls is not None:
df = star.select_classes(df, args.cls)
star.set_original_fields(df, inplace=True)
if args.resort:
df = df.sort_values([star.UCSF.IMAGE_ORIGINAL_PATH,
star.UCSF.IMAGE_ORIGINAL_INDEX])
for idx, row in df.iterrows():
if args.stack_path is not None:
input_stack_path = os.path.join(args.stack_path, row[star.UCSF.IMAGE_ORIGINAL_PATH])
else:
input_stack_path = row[star.UCSF.IMAGE_ORIGINAL_PATH]
with mrc.ZSliceReader(input_stack_path) as reader:
i = row[star.UCSF.IMAGE_ORIGINAL_INDEX]
writer.write(reader.read(i))
elif fn.endswith(".par"):
if args.stack_path is None:
log.error(".par file input requires --stack-path")
return 1
df = metadata.par2star(metadata.parse_fx_par(fn), data_path=args.stack_path)
# star.set_original_fields(df, inplace=True) # Redundant.
star.augment_star_ucsf(df)
elif fn.endswith(".csv"):
return 1
elif fn.endswith(".cs"):
return 1
else:
if fn.endswith(".mrcs"):
with mrc.ZSliceReader(fn) as reader:
for img in reader:
writer.write(img)
df = pd.DataFrame(
{star.UCSF.IMAGE_ORIGINAL_INDEX: np.arange(reader.nz)})
df[star.UCSF.IMAGE_ORIGINAL_PATH] = fn
else:
print("Unrecognized input file type")
return 1
if args.star is not None:
df[star.UCSF.IMAGE_INDEX] = np.arange(first_ptcl,
first_ptcl + df.shape[0])
if args.abs_path:
df[star.UCSF.IMAGE_PATH] = writer.path
else:
df[star.UCSF.IMAGE_PATH] = os.path.relpath(writer.path, os.path.dirname(args.star))
df["index"] = df[star.UCSF.IMAGE_INDEX]
star.simplify_star_ucsf(df)
dfs.append(df)
first_ptcl += df.shape[0]
if args.star is not None:
df = pd.concat(dfs, join="inner")
# df = pd.concat(dfs)
# df = df.dropna(df, axis=1, how="any")
if not args.relion2: # Relion 3.1 style output.
df = star.remove_deprecated_relion2(df, inplace=True)
star.write_star(args.star, df, resort_records=False, optics=True)
else:
df = star.remove_new_relion31(df, inplace=True)
star.write_star(args.star, df, resort_records=False, optics=False)
return 0