447 lines
16 KiB
Python
Executable File
447 lines
16 KiB
Python
Executable File
#!/usr/bin/python
|
|
|
|
# author J.Beale
|
|
|
|
"""
|
|
# aim
|
|
crystfel/0.12.0 variant of crystfel_split.py for testing the Millepede-II
|
|
detector geometry refinement pipeline (align_detector) as a replacement for
|
|
the old crystfel/0.10.2 detector-shift workflow.
|
|
|
|
to process a batch of data very fast by splitting it into a number of chunks and submitting
|
|
these jobs separately to the cluster
|
|
|
|
# usage
|
|
python crystfel_split_mille.py -l <path-to-list-file>
|
|
-k <chunk-size>
|
|
-g <path-to-geom-file>
|
|
-c <path-to-cell-file>
|
|
-n <name-of-job>
|
|
-p photons or
|
|
|
|
# crystfel parameter may need some editing in the function - write_crystfel_run
|
|
# the geometry file must use the new "group" hierarchy syntax (group_all = ...)
|
|
# rather than the legacy rigid_group_* syntax -- see JF17T16V01-millepede.geom
|
|
|
|
# output
|
|
a series of stream files from crystfel in the current working directory
|
|
a concatenated .stream file in cwd
|
|
a series of Millepede-II binary files (<name>_<chunk>-mille.dat) per chunk
|
|
a plain-text list of those mille file paths (<name>-mille.lst), for feeding
|
|
straight into align_detector
|
|
a log file with .geom and evalation of indexing, cell etc
|
|
"""
|
|
|
|
# modules
|
|
import argparse
|
|
import pandas as pd
|
|
import subprocess
|
|
import os, errno
|
|
import time
|
|
from tqdm import tqdm
|
|
import regex as re
|
|
import numpy as np
|
|
from loguru import logger
|
|
|
|
def scrub_cells( line ):
|
|
|
|
# get uc values from stream file
|
|
# example - Cell parameters 7.71784 7.78870 3.75250 nm, 90.19135 90.77553 90.19243 deg
|
|
pattern = r"Cell\sparameters\s(\d+\.\d+)\s(\d+\.\d+)\s(\d+\.\d+)\snm,\s(\d+\.\d+)\s(\d+\.\d+)\s(\d+\.\d+)\sdeg"
|
|
a = re.search( pattern, line ).group(1)
|
|
b = re.search( pattern, line ).group(2)
|
|
c = re.search( pattern, line ).group(3)
|
|
alpha = re.search( pattern, line ).group(4)
|
|
beta = re.search( pattern, line ).group(5)
|
|
gamma = re.search( pattern, line ).group(6)
|
|
|
|
return [ a, b, c, alpha, beta, gamma ]
|
|
|
|
def scrub_res( stream ):
|
|
|
|
# get diffraction limit
|
|
# example - diffraction_resolution_limit = 4.07 nm^-1 or 2.46 A
|
|
pattern = r"diffraction_resolution_limit\s=\s\d\.\d+\snm\^-1\sor\s(\d+\.\d+)\sA"
|
|
res = re.search( pattern, stream ).group(1)
|
|
return res
|
|
|
|
def scrub_obs( stream ):
|
|
|
|
# get number of reflections
|
|
# example - num_reflections = 308
|
|
pattern = r"num_reflections\s=\s(\d+)"
|
|
obs = re.search( pattern, stream ).group(1)
|
|
return obs
|
|
|
|
def calculate_stats( stream_pwd ):
|
|
|
|
chunks = 0
|
|
xtals = 0
|
|
cells = []
|
|
obs_list = []
|
|
res_list = []
|
|
|
|
print( "scrubing data" )
|
|
# open stream file
|
|
with open( stream_pwd ) as stream:
|
|
for line in stream:
|
|
|
|
# count chunks
|
|
if line.startswith( "----- Begin chunk -----" ):
|
|
chunks = chunks + 1
|
|
|
|
# get cell
|
|
if line.startswith( "Cell parameters" ):
|
|
cell = scrub_cells( line )
|
|
cells.append( cell )
|
|
xtals = xtals + 1
|
|
|
|
# get res
|
|
if line.startswith( "diffraction_resolution_limit" ):
|
|
res = scrub_res( line )
|
|
res_list.append( res )
|
|
|
|
# get obs
|
|
if line.startswith( "num_reflections" ):
|
|
obs = scrub_obs( line )
|
|
obs_list.append( obs )
|
|
|
|
if chunks % 1000 == 0:
|
|
print( "scrubbed {0} chunks".format( chunks ), end='\r' )
|
|
|
|
# res_df
|
|
cols = [ "a", "b", "c", "alpha", "beta", "gamma" ]
|
|
df = pd.DataFrame( cells, columns=cols )
|
|
df[ "resolution" ] = res_list
|
|
df[ "obs" ] = obs_list
|
|
|
|
# convert all to floats
|
|
df = df.astype(float)
|
|
|
|
return df, xtals, chunks
|
|
|
|
def h5_split( lst, chunk_size ):
|
|
|
|
# read h5.lst - note - removes // from image column
|
|
# scrub file name
|
|
lst_name = os.path.basename( lst )
|
|
|
|
cols = [ "h5", "image" ]
|
|
df = pd.read_csv( lst, sep="\s//", engine="python", names=cols )
|
|
|
|
# re-add // to image columm and drop other columns
|
|
df[ "h5_path" ] = df.h5 + " //" + df.image.astype( str )
|
|
df = df[ [ "h5_path" ] ]
|
|
|
|
# split df into a lst
|
|
list_df = [df[i:i + chunk_size] for i in range( 0, len(df), chunk_size)]
|
|
|
|
return list_df
|
|
|
|
def write_crystfel_run( proc_dir, name, chunk, chunk_lst_file, geom_file, cell_file, threshold ):
|
|
|
|
"""
|
|
crystfel run file - spot-finding and indexing parameters may need some editing
|
|
only change from inside the quote ("")
|
|
"""
|
|
|
|
# stream file name
|
|
stream_file = "{0}_{1}.stream".format( name, chunk )
|
|
|
|
# millepede-II binary output for this chunk (do NOT concatenate these -
|
|
# they are binary and get passed to align_detector as a file list)
|
|
mille_file = "{0}_{1}-mille.dat".format( name, chunk )
|
|
|
|
# crystfel file name
|
|
cryst_run_file = "{0}/{1}_{2}.sh".format( proc_dir, name, chunk )
|
|
# write file
|
|
run_sh = open( cryst_run_file, "w" )
|
|
run_sh.write( "#!/bin/sh\n\n" )
|
|
run_sh.write( "module purge\n" )
|
|
run_sh.write( "module use MX unstable\n" )
|
|
run_sh.write( "module load crystfel/0.12.0\n" )
|
|
run_sh.write( "indexamajig -i {0} \\\n".format( chunk_lst_file ) )
|
|
run_sh.write( " --output={0} \\\n".format( stream_file ) )
|
|
run_sh.write( " --geometry={0} \\\n".format( geom_file ) )
|
|
run_sh.write( " --pdb={0} \\\n".format( cell_file ) )
|
|
run_sh.write( " --push-res=0.5 \\\n" )
|
|
run_sh.write( " --indexing=xgandalf-latt-cell \\\n" )
|
|
run_sh.write( " --peaks=peakfinder8 \\\n" )
|
|
run_sh.write( " --integration=rings-nocen-nograd \\\n" )
|
|
run_sh.write( " --tolerance=10.0,10.0,10.0,2,3,2 \\\n" )
|
|
run_sh.write( " --threshold={0} \\\n".format( threshold ) )
|
|
run_sh.write( " --min-snr=5 \\\n" )
|
|
run_sh.write( " --int-radius=4,6,7 \\\n" )
|
|
run_sh.write( " -j 32 \\\n" )
|
|
run_sh.write( " --multi \\\n" )
|
|
run_sh.write( " --check-peaks \\\n" )
|
|
run_sh.write( " --retry \\\n" )
|
|
run_sh.write( " --max-res=3000 \\\n" )
|
|
run_sh.write( " --min-pix-count=2 \\\n" )
|
|
run_sh.write( " --local-bg-radius=4 \\\n" )
|
|
run_sh.write( " --min-res=85 \\\n" )
|
|
run_sh.write( " --mille \\\n" )
|
|
run_sh.write( " --mille-file={0}".format( mille_file ) )
|
|
run_sh.close()
|
|
|
|
# make file executable
|
|
subprocess.call( [ "chmod", "+x", "{0}".format( cryst_run_file ) ] )
|
|
|
|
# return crystfel file name
|
|
return cryst_run_file, stream_file, mille_file
|
|
|
|
def make_process_dir( proc_dir ):
|
|
# make process directory
|
|
try:
|
|
os.makedirs( proc_dir )
|
|
except OSError as e:
|
|
if e.errno != errno.EEXIST:
|
|
logger.debug( "making directory error" )
|
|
raise
|
|
|
|
def submit_job( job_file, reservation ):
|
|
|
|
# submit the job
|
|
if reservation:
|
|
print( "using a ra beamtime reservation = {0}".format( reservation ) )
|
|
logger.info( "using ra reservation to process data = {0}".format( reservation ) )
|
|
submit_cmd = [ "sbatch", "-p", "hour", "--reservation={0}".format( reservation ), "--cpus-per-task=32", "--" , job_file ]
|
|
else:
|
|
submit_cmd = [ "sbatch", "-p", "hour", "--cpus-per-task=32", "--" , job_file ]
|
|
logger.info( "using slurm command = {0}".format( submit_cmd ) )
|
|
|
|
try:
|
|
job_output = subprocess.check_output( submit_cmd )
|
|
logger.info( "submited job = {0}".format( job_output ) )
|
|
except subprocess.CalledProcessError as e:
|
|
print( "please give the correct ra reservation or remove the -v from the arguements" )
|
|
exit()
|
|
|
|
# scrub job id from - example Submitted batch job 742403
|
|
pattern = r"Submitted batch job (\d+)"
|
|
job_id = re.search( pattern, job_output.decode().strip() ).group(1)
|
|
|
|
return int( job_id )
|
|
|
|
def wait_for_jobs( job_ids, total_jobs ):
|
|
|
|
with tqdm( total=total_jobs, desc="Jobs Completed", unit="job" ) as pbar:
|
|
while job_ids:
|
|
completed_jobs = set()
|
|
for job_id in job_ids:
|
|
status_cmd = [ "squeue", "-h", "-j", str( job_id ) ]
|
|
status = subprocess.check_output(status_cmd)
|
|
if not status:
|
|
completed_jobs.add(job_id)
|
|
pbar.update(1)
|
|
job_ids.difference_update(completed_jobs)
|
|
time.sleep(2)
|
|
|
|
def run_splits( list_df, cwd, name, geom_file, cell_file, threshold, reservation ):
|
|
|
|
# set chunk counter
|
|
chunk = 0
|
|
|
|
# submitted job set
|
|
submitted_job_ids = set()
|
|
|
|
# stream and mille file lists
|
|
stream_lst = []
|
|
mille_lst = []
|
|
|
|
for chunk_lst in list_df:
|
|
|
|
logger.info( "chunk {0} = {1} images".format( chunk, len( chunk_lst ) ) )
|
|
# define process directory
|
|
proc_dir = "{0}/{1}/{1}_{2}".format( cwd, name, chunk )
|
|
|
|
# make process directory
|
|
make_process_dir(proc_dir)
|
|
|
|
# move to process directory
|
|
os.chdir( proc_dir )
|
|
|
|
# write list to file
|
|
chunk_lst_file = "{0}/{1}_{2}.lst".format( proc_dir, name, chunk )
|
|
chunk_lst.to_csv( chunk_lst_file, index=False, header=False )
|
|
|
|
# write crystfel file and append paths to lists
|
|
cryst_run_file, stream_file, mille_file = write_crystfel_run( proc_dir, name, chunk, chunk_lst_file, geom_file, cell_file, threshold )
|
|
stream_lst.append( "{0}/{1}".format( proc_dir, stream_file ) )
|
|
mille_lst.append( "{0}/{1}".format( proc_dir, mille_file ) )
|
|
|
|
# submit jobs
|
|
job_id = submit_job( cryst_run_file, reservation )
|
|
submitted_job_ids.add( job_id )
|
|
|
|
# increase chunk counter
|
|
chunk = chunk +1
|
|
|
|
# move back to top dir
|
|
os.chdir( cwd )
|
|
|
|
return submitted_job_ids, chunk, stream_lst, mille_lst
|
|
|
|
def main( cwd, name, lst, chunk_size, geom_file, cell_file, threshold, reservation ):
|
|
|
|
print( "reading SwissFEL lst file" )
|
|
print( "creating {0} image chunks of lst".format( chunk_size ) )
|
|
list_df = h5_split( lst, chunk_size )
|
|
print( "DONE" )
|
|
|
|
# run crystfel runs on individual splits
|
|
print( "submitting jobs to cluster" )
|
|
submitted_job_ids, chunk, stream_lst, mille_lst = run_splits( list_df, cwd, name, geom_file, cell_file, threshold, reservation )
|
|
|
|
# monitor progress of jobs
|
|
time.sleep( 30 )
|
|
wait_for_jobs( submitted_job_ids, chunk )
|
|
print( "done" )
|
|
|
|
# make composite .stream file
|
|
output_file = "{0}.stream".format( name )
|
|
|
|
print( "concatenating .streams from separate runs." )
|
|
try:
|
|
# Open the output file in 'append' mode
|
|
with open(output_file, "a") as output:
|
|
for file_name in stream_lst:
|
|
try:
|
|
with open(file_name, "r") as input_file:
|
|
# Read the contents of the input file and append to the output file
|
|
output.write(input_file.read())
|
|
except FileNotFoundError:
|
|
logger.debug(f"File {file_name} not found. Skipping.")
|
|
except IOError as e:
|
|
logger.debug(f"An error occurred while appending files: {e}")
|
|
|
|
print( "done" )
|
|
|
|
# write out the list of mille files (NOT concatenated - binary format,
|
|
# align_detector takes them as separate positional arguments)
|
|
mille_list_file = "{0}-mille.lst".format( name )
|
|
print( "writing mille file list to {0}".format( mille_list_file ) )
|
|
with open( mille_list_file, "w" ) as f:
|
|
for file_name in mille_lst:
|
|
if os.path.exists( file_name ):
|
|
f.write( file_name + "\n" )
|
|
else:
|
|
logger.debug(f"Mille file {file_name} not found. Skipping.")
|
|
|
|
df, xtals, chunks = calculate_stats( output_file )
|
|
|
|
# stats
|
|
index_rate = round( xtals/chunks*100, 2 )
|
|
mean_res, std_res = round( df.resolution.mean(), 2 ), round( df.resolution.std(), 2 )
|
|
median_res = df.resolution.median()
|
|
mean_obs, std_obs = round( df.obs.mean(), 2 ), round( df.obs.std(), 2)
|
|
mean_a, std_a = round( df.a.mean()*10, 2 ), round( df.a.std()*10, 2 )
|
|
mean_b, std_b = round( df.b.mean()*10, 2 ), round( df.b.std()*10, 2 )
|
|
mean_c, std_c = round( df.c.mean()*10, 2 ), round( df.c.std()*10, 2 )
|
|
mean_alpha, std_alpha = round( df.alpha.mean(), 2 ), round( df.alpha.std(), 2 )
|
|
mean_beta, std_beta = round(df.beta.mean(), 2 ), round( df.beta.std(), 2 )
|
|
mean_gamma, std_gamma = round( df.gamma.mean(), 2 ), round( df.gamma.std(), 2 )
|
|
|
|
logger.info( "images = {0}".format( chunks ) )
|
|
logger.info( "crystals = {0}".format( xtals ) )
|
|
logger.info( "indexing rate = {0} %".format( index_rate ) )
|
|
logger.info( "mean resolution = {0} +/- {1} A".format( mean_res, std_res ) )
|
|
logger.info( "median resolution = {0} A".format( median_res ) )
|
|
logger.info( "mean observations = {0} +/- {1}".format( mean_obs, std_obs ) )
|
|
logger.info( "mean a = {0} +/- {1} A".format( mean_a, std_a ) )
|
|
logger.info( "mean b = {0} +/- {1} A".format( mean_b, std_b ) )
|
|
logger.info( "mean c = {0} +/- {1} A".format( mean_c, std_c ) )
|
|
logger.info( "mean alpha = {0} +/- {1} deg".format( mean_alpha, std_alpha ) )
|
|
logger.info( "mean beta = {0} +/- {1} deg".format( mean_beta, std_beta ) )
|
|
logger.info( "mean gamma = {0} +/- {1} deg".format( mean_gamma, std_gamma ) )
|
|
|
|
print( "printing stats" )
|
|
print( "images = {0}".format( chunks ) )
|
|
print( "crystals = {0}".format( xtals ) )
|
|
print( "indexing rate = {0} %".format( index_rate ) )
|
|
print( "mean resolution = {0} +/- {1} A".format( mean_res, std_res ) )
|
|
print( "median resolution = {0} A".format( median_res ) )
|
|
print( "mean observations = {0} +/- {1}".format( mean_obs, std_obs ) )
|
|
print( "mean a = {0} +/- {1} A".format( mean_a, std_a ) )
|
|
print( "mean b = {0} +/- {1} A".format( mean_b, std_b ) )
|
|
print( "mean c = {0} +/- {1} A".format( mean_c, std_c ) )
|
|
print( "mean alpha = {0} +/- {1} deg".format( mean_alpha, std_alpha ) )
|
|
print( "mean beta = {0} +/- {1} deg".format( mean_beta, std_beta ) )
|
|
print( "mean gamma = {0} +/- {1} deg".format( mean_gamma, std_gamma ) )
|
|
|
|
if __name__ == "__main__":
|
|
parser = argparse.ArgumentParser()
|
|
parser.add_argument(
|
|
"-l",
|
|
"--lst_file",
|
|
help="file from SwissFEL output to be processed quickly. Requried.",
|
|
type=os.path.abspath,
|
|
required=True
|
|
)
|
|
parser.add_argument(
|
|
"-k",
|
|
"--chunk_size",
|
|
help="how big should each chunk be? - the bigger the chunk, the fewer jobs, the slower it will be. Default = 500.",
|
|
type=int,
|
|
default=500
|
|
)
|
|
parser.add_argument(
|
|
"-g",
|
|
"--geom_file",
|
|
help="path to geom file to be used in the refinement. Must use the new 'group' hierarchy syntax (see JF17T16V01-millepede.geom). Required.",
|
|
type=os.path.abspath,
|
|
required=True
|
|
)
|
|
parser.add_argument(
|
|
"-c",
|
|
"--cell_file",
|
|
help="path to cell file of the crystals used in the refinement. Required.",
|
|
type=os.path.abspath,
|
|
required=True
|
|
)
|
|
parser.add_argument(
|
|
"-n",
|
|
"--job_name",
|
|
help="the name of the job to be done. Default = 'split_###'",
|
|
type=str,
|
|
default="split"
|
|
)
|
|
parser.add_argument(
|
|
"-v",
|
|
"--reservation",
|
|
help="reservation name for ra cluster. Usually along the lines of P11111_2024-12-10",
|
|
type=str,
|
|
default=None
|
|
)
|
|
parser.add_argument(
|
|
"-p",
|
|
"--photons_or_energy",
|
|
help="determines the threshold to use for CrystFEL. Photons counts have always been used in Cristallina and are now used on Alvra from 01.11.2024. Please use 'energy' for Alvra before this.",
|
|
type=str,
|
|
default="photons"
|
|
)
|
|
parser.add_argument(
|
|
"-d",
|
|
"--debug",
|
|
help="output debug to terminal.",
|
|
type=bool,
|
|
default=False
|
|
)
|
|
args = parser.parse_args()
|
|
# set current working directory
|
|
cwd = os.getcwd()
|
|
# set loguru
|
|
if not args.debug:
|
|
logger.remove()
|
|
logfile = "{0}.log".format( args.job_name )
|
|
logger.add( logfile, format="{message}", level="INFO")
|
|
# log geometry file
|
|
geom = open( args.geom_file, "r" ).read()
|
|
logger.info( geom )
|
|
# set threshold based on detector
|
|
if args.photons_or_energy == "energy":
|
|
threshold = 3000
|
|
elif args.photons_or_energy == "photons":
|
|
threshold = 15
|
|
main( cwd, args.job_name, args.lst_file, args.chunk_size, args.geom_file, args.cell_file, threshold, args.reservation )
|