|
|
[Rivet-svn] r3081 - trunk/bin
blackhole at projects.hepforge.org
blackhole at projects.hepforge.org
Thu May 5 18:27:38 BST 2011
Author: hoeth
Date: Thu May 5 18:27:38 2011
New Revision: 3081
Log:
next try to get rid of the duplicates in compare-histos
Modified:
trunk/bin/compare-histos
Modified: trunk/bin/compare-histos
==============================================================================
--- trunk/bin/compare-histos Thu May 5 17:38:00 2011 (r3080)
+++ trunk/bin/compare-histos Thu May 5 18:27:38 2011 (r3081)
@@ -46,38 +46,60 @@
sys.exit(1)
+## Function to make output dirs
+def mkoutdir(outdir):
+ if not os.path.exists(outdir):
+ try:
+ os.makedirs(outdir)
+ except:
+ msg = "Can't make output directory '%s'" % outdir
+ logging.error(msg)
+ raise Exception(msg)
+ if not os.access(outdir, os.W_OK):
+ msg = "Can't write to output directory '%s'" % outdir
+ logging.error(msg)
+ raise Exception(msg)
+
+
+def getHistos(aidafile):
+ '''Get a dictionary of histograms indexed by name.'''
+ if not re.match(r'.*\.aida$', aidafile):
+ logging.error("Error: input file '%s' is not an AIDA file" % aidafile)
+ sys.exit(2)
+ aidafilepath = os.path.abspath(aidafile)
+ if not os.access(aidafilepath, os.R_OK):
+ logging.error("Error: cannot read from %s" % aidafile)
+ sys.exit(2)
+
+ histos, titles, xlabels, ylabels = {}, {}, {}, {}
+ tree = ET.parse(aidafilepath)
+ for dps in tree.findall("dataPointSet"):
+ ## Get this histogram's path name
+ dpsname = os.path.join(dps.get("path"), dps.get("name"))
+ h = Histo.fromDPS(dps)
+ ## Is it a data histo?
+ h.isdata = dpsname.upper().startswith("/REF")
+ if h.isdata:
+ dpsname = dpsname.replace("/REF", "")
+ if not titles.has_key(dpsname):
+ titles[dpsname] = h.title
+ xlabels[dpsname] = h.xlabel
+ ylabels[dpsname] = h.ylabel
+ else:
+ if dpsname.count('/') > 2:
+ dpsname = '/' + dpsname.split('/', 2)[-1]
+ titles[dpsname] = h.title
+ xlabels[dpsname] = h.xlabel
+ ylabels[dpsname] = h.ylabel
+ h.expt = dpsname.split("_")[0][1:]
+ ## Hard-coded cosmetic handling for the D0 experiment name!
+ if h.expt == "D0":
+ h.expt = "D\O\ "
+ histos[dpsname] = h
+ return histos, titles, xlabels, ylabels
-if __name__ == "__main__":
- import os, re, logging
-
- PROGPATH = sys.argv[0]
- PROGNAME = os.path.basename(PROGPATH)
-
- ## Try to rename the process on Linux
- try:
- import ctypes
- libc = ctypes.cdll.LoadLibrary('libc.so.6')
- libc.prctl(15, 'compare-histos', 0, 0, 0)
- except Exception:
- pass
-
- ## Try to use Psyco optimiser
- try:
- import psyco
- psyco.full()
- except ImportError:
- pass
-
- ## Get Rivet data dir
- rivet_data_dirs = None
- try:
- import rivet
- rivet_data_dirs = rivet.getAnalysisRefPaths()
- except Exception, e:
- sys.stderr.write(PROGNAME + " requires the 'rivet' Python module\n")
- logging.debug(str(e))
- sys.exit(1)
+def getCommandLineOptions():
## Parse command line options
from optparse import OptionParser, OptionGroup
parser = OptionParser(usage=__doc__)
@@ -138,7 +160,43 @@
verbgrp.add_option("-v", "--verbose", help="Add extra debug messages", dest="LOGLEVEL",
action="store_const", default=logging.INFO, const=logging.DEBUG)
parser.add_option_group(verbgrp)
+ return parser
+
+
+##################################################################
+
+if __name__ == "__main__":
+ import os, re, logging
+
+ PROGPATH = sys.argv[0]
+ PROGNAME = os.path.basename(PROGPATH)
+
+ ## Try to rename the process on Linux
+ try:
+ import ctypes
+ libc = ctypes.cdll.LoadLibrary('libc.so.6')
+ libc.prctl(15, 'compare-histos', 0, 0, 0)
+ except Exception:
+ pass
+ ## Try to use Psyco optimiser
+ try:
+ import psyco
+ psyco.full()
+ except ImportError:
+ pass
+
+ ## Get Rivet data dir
+ rivet_data_dirs = None
+ try:
+ import rivet
+ rivet_data_dirs = rivet.getAnalysisRefPaths()
+ except Exception, e:
+ sys.stderr.write(PROGNAME + " requires the 'rivet' Python module\n")
+ logging.debug(str(e))
+ sys.exit(1)
+
+ parser = getCommandLineOptions()
opts, args = parser.parse_args()
@@ -187,15 +245,15 @@
for ls in LINESTYLES:
STYLES.append( (c, ls) )
+
## Get file names and labels
FILES = []
- FILEOPTIONS = { }
+ REFFILES = []
+ FILEOPTIONS = {}
if opts.RIVETREFS and rivet_data_dirs:
- reffiles = []
for d in rivet_data_dirs:
import glob
- reffiles += glob.glob(os.path.join(d, "*.aida"))
- args = reffiles + args
+ REFFILES += glob.glob(os.path.join(d, "*.aida"))
for a in args:
asplit = a.split(":")
path = asplit[0]
@@ -207,6 +265,8 @@
asplit[i] = "Title=%s" % asplit[i]
FILEOPTIONS[path].append(asplit[i])
+ ## Ignore duplicates
+ REFFILES = list(set(REFFILES))
## Check that the requested files are sensible
if (len(FILES) < 1):
@@ -221,44 +281,6 @@
sys.exit(2)
- def getHistos(aidafile):
- '''Get a dictionary of histograms indexed by name.'''
- if not re.match(r'.*\.aida$', aidafile):
- logging.error("Error: input file '%s' is not an AIDA file" % aidafile)
- sys.exit(2)
- aidafilepath = os.path.abspath(aidafile)
- if not os.access(aidafilepath, os.R_OK):
- logging.error("Error: cannot read from %s" % aidafile)
- sys.exit(2)
-
- histos, titles, xlabels, ylabels = {}, {}, {}, {}
- tree = ET.parse(aidafilepath)
- for dps in tree.findall("dataPointSet"):
- ## Get this histogram's path name
- dpsname = os.path.join(dps.get("path"), dps.get("name"))
- h = Histo.fromDPS(dps)
- ## Is it a data histo?
- h.isdata = dpsname.upper().startswith("/REF")
- if h.isdata:
- dpsname = dpsname.replace("/REF", "")
- if not titles.has_key(dpsname):
- titles[dpsname] = h.title
- xlabels[dpsname] = h.xlabel
- ylabels[dpsname] = h.ylabel
- else:
- if dpsname.count('/') > 2:
- dpsname = '/' + dpsname.split('/', 2)[-1]
- titles[dpsname] = h.title
- xlabels[dpsname] = h.xlabel
- ylabels[dpsname] = h.ylabel
- h.expt = dpsname.split("_")[0][1:]
- ## Hard-coded cosmetic handling for the D0 experiment name!
- if h.expt == "D0":
- h.expt = "D\O\ "
- histos[dpsname] = h
- return histos, titles, xlabels, ylabels
-
-
## Read histo data from files into data structures
HISTOS = {}
TITLES = {}
@@ -267,10 +289,10 @@
LABELS = {}
NAMES = set()
MCNAMES = set()
- for f in FILES:
+ for f in FILES+REFFILES:
HISTOS[f] = {}
LABELS[f] = {}
- for f in FILES:
+ for f in FILES+REFFILES:
histos, titles, xlabels, ylabels = getHistos(f)
for n, h in histos.iteritems():
if h.isdata:
@@ -313,20 +335,6 @@
MCNAMES = MCNAMES.intersection(hnames)
- ## Function to make output dirs
- def mkoutdir(outdir):
- if not os.path.exists(outdir):
- try:
- os.makedirs(outdir)
- except:
- msg = "Can't make output directory '%s'" % outdir
- logging.error(msg)
- raise Exception(msg)
- if not os.access(outdir, os.W_OK):
- msg = "Can't write to output directory '%s'" % outdir
- logging.error(msg)
- raise Exception(msg)
-
## Pre-emptively reduce number of files to iterate through
## (assuming, reasonably, that there is only one ref file per histo)
activenames = NAMES
@@ -351,6 +359,12 @@
## Identify contributing data files for this histo
activemcfiles = []
activereffiles = []
+ for f in REFFILES:
+ if HISTOS.has_key(f):
+ d = HISTOS[f]
+ if d.has_key(name):
+ if d[name].isdata:
+ activereffiles.append(f)
for f in FILES:
if HISTOS.has_key(f):
d = HISTOS[f]
@@ -361,6 +375,7 @@
activemcfiles.append(f)
activefiles = activereffiles + activemcfiles
#print activereffiles
+ #print activemcfiles
#print activefiles
if len(activefiles) == 0:
logging.warning("Something's wrong... somehow there's no data for histogram '%s'!" % name)
More information about the Rivet-svn
mailing list
|