_exit(usage, 1, e)
doer = DU_ABPTable(sModelName,
sModelDir,
C=options.crf_C,
tol=options.crf_tol,
njobs=options.crf_njobs,
max_iter=options.max_iter,
inference_cache=options.crf_inference_cache)
if options.rm:
doer.rm()
sys.exit(0)
lTrn, lTst, lRun, lFold = [
_checkFindColDir(lsDir)
for lsDir in [options.lTrn, options.lTst, options.lRun, options.lFold]
]
# if options.bAnnotate:
# doer.annotateDocument(lTrn)
# traceln('annotation done')
# sys.exit(0)
traceln("- classes: ", doer.getGraphClass().getLabelNameList())
## use. a_mpxml files
doer.sXmlFilenamePattern = doer.sLabeledXmlFilenamePattern
if options.iFoldInitNum or options.iFoldRunNum or options.bFoldFinish:
if options.iFoldInitNum:
"""
# ---
#parse the command line
(options, args) = parser.parse_args()
# ---
try:
sModelDir, sModelName = args
except Exception as e:
traceln("Specify a model folder and a model name!")
_exit(usage, 1, e)
doer = DU_ABPTableAnnotator(sModelName, sModelDir,
C = options.crf_C,
tol = options.crf_tol,
njobs = options.crf_njobs,
max_iter = options.crf_max_iter,
inference_cache = options.crf_inference_cache)
if options.rm:
doer.rm()
sys.exit(0)
lTrn, lTst, lRun, lFold = [_checkFindColDir(lsDir) for lsDir in [options.lTrn, options.lTst, options.lRun, options.lFold]]
doer.annotateDocument(lTrn)
traceln('annotation done')
inference_cache=options.crf_inference_cache)
if options.rm:
doer.rm()
sys.exit(0)
traceln("- classes: ", DU_GRAPH.getLabelNameList())
if options.best_params:
dBestParams = doer.getModelClass().loadBestParams(
sModelDir, options.best_params)
doer.setLearnerConfiguration(dBestParams)
#Add the "out" subdir if needed
lTrn, lTst, lRun, lFold = [
_checkFindColDir(lsDir, "out")
for lsDir in [options.lTrn, options.lTst, options.lRun, options.lFold]
]
if options.iFoldInitNum or options.iFoldRunNum or options.bFoldFinish:
if options.iFoldInitNum:
"""
initialization of a cross-validation
"""
splitter, ts_trn, lFilename_trn = doer._nfold_Init(
lFold, options.iFoldInitNum, bStoreOnDisk=True)
elif options.iFoldRunNum:
"""
Run one fold
"""
oReport = doer._nfold_RunFoldFromDisk(options.iFoldRunNum,
def main(sModelDir, sModelName, options):
doer = DU_ABPTableR(sModelName, sModelDir,
C = options.crf_C,
tol = options.crf_tol,
njobs = options.crf_njobs,
max_iter = options.max_iter,
inference_cache = options.crf_inference_cache)
if options.rm:
doer.rm()
return
lTrn, lTst, lRun, lFold = [_checkFindColDir(lsDir, bAbsolute=False) for lsDir in [options.lTrn, options.lTst, options.lRun, options.lFold]]
# if options.bAnnotate:
# doer.annotateDocument(lTrn)
# traceln('annotation done')
# sys.exit(0)
traceln("- classes: ", doer.getGraphClass().getLabelNameList())
## use. a_mpxml files
#doer.sXmlFilenamePattern = doer.sLabeledXmlFilenamePattern
if options.iFoldInitNum or options.iFoldRunNum or options.bFoldFinish:
if options.iFoldInitNum:
"""
initialization of a cross-validation
"""
splitter, ts_trn, lFilename_trn = doer._nfold_Init(lFold, options.iFoldInitNum, bStoreOnDisk=True)
elif options.iFoldRunNum:
"""
Run one fold
"""
oReport = doer._nfold_RunFoldFromDisk(options.iFoldRunNum, options.warm, options.pkl)
traceln(oReport)
elif options.bFoldFinish:
tstReport = doer._nfold_Finish()
traceln(tstReport)
else:
assert False, "Internal error"
#no more processing!!
exit(0)
#-------------------
if lFold:
loTstRpt = doer.nfold_Eval(lFold, 3, .25, None, options.pkl)
sReportPickleFilename = os.path.join(sModelDir, sModelName + "__report.txt")
traceln("Results are in %s"%sReportPickleFilename)
graph.GraphModel.GraphModel.gzip_cPickle_dump(sReportPickleFilename, loTstRpt)
elif lTrn:
doer.train_save_test(lTrn, lTst, options.warm, options.pkl)
try: traceln("Baseline best estimator: %s"%doer.bsln_mdl.best_params_) #for GridSearch
except: pass
traceln(" --- CRF Model ---")
traceln(doer.getModel().getModelInfo())
elif lTst:
doer.load()
tstReport = doer.test(lTst)
traceln(tstReport)
if options.bDetailedReport:
traceln(tstReport.getDetailledReport())
sReportPickleFilename = os.path.join(sModelDir, sModelName + "__detailled_report.txt")
graph.GraphModel.GraphModel.gzip_cPickle_dump(sReportPickleFilename, tstReport)
if lRun:
if options.storeX or options.applyY:
try: doer.load()
except: pass #we only need the transformer
lsOutputFilename = doer.runForExternalMLMethod(lRun, options.storeX, options.applyY, options.bRevertEdges)
else:
doer.load()
lsOutputFilename = doer.predict(lRun)
traceln("Done, see in:\n %s"%lsOutputFilename)
def test_test_ensemble(self):
fname = os.path.join('UT_MODELS', 'UT_mod2_m2_bestmodel.ckpt.index')
model_dir = 'UT_MODELS'
model_name = 'UT_mod2_ensemble'
lTrn = _checkFindColDir('./abp_test')
dLearnerConfig = \
{
"ecn_ensemble": [
{
"type": "ecn",
"name": "UT_mod2_m1",
"dropout_rate_edge": 0.0,
"dropout_rate_edge_feat": 0.0,
"dropout_rate_node": 0.0,
"lr": 0.001,
"mu": 0.0,
"nb_iter": 50,
"nconv_edge": 1,
"node_indim": -1,
"num_layers": 8,
"ratio_train_val": 0.1,
"patience": 500,
"activation_name": "tanh",
"stack_convolutions": True
}
,
{
"type": "ecn",
"name": "UT_mod2_m2",
"dropout_rate_edge": 0.0,
"dropout_rate_edge_feat": 0.0,
"dropout_rate_node": 0.0,
"lr": 0.001,
"mu": 0.0,
"nb_iter": 50,
"nconv_edge": 1,
"node_indim": -1,
"num_layers": 8,
"ratio_train_val": 0.1,
"patience": 50,
"activation_name": "relu",
"stack_convolutions": True
}
]
}
if os.path.exists(fname):
#Do the test
doer = DU_ABPTable.DU_ABPTable_ECN(
model_name, model_dir, dLearnerConfigArg=dLearnerConfig)
doer.load()
tstReport = doer.test(lTrn)
acc, _ = tstReport.getClassificationReport()
print(acc)
self.assertTrue(acc > 0.5)
else:
self.fail('UT_mod1 was not trained')
def main(sModelDir, sModelName, options):
doer = DU_ABPTable_TypedCRF(sModelName,
sModelDir,
C=options.crf_C,
tol=options.crf_tol,
njobs=options.crf_njobs,
max_iter=options.max_iter,
inference_cache=options.crf_inference_cache)
if options.docid:
sDocId = options.docid
else:
sDocId = None
if options.rm:
doer.rm()
return
lTrn, lTst, lRun, lFold = [
_checkFindColDir(lsDir)
for lsDir in [options.lTrn, options.lTst, options.lRun, options.lFold]
]
# if options.bAnnotate:
# doer.annotateDocument(lTrn)
# traceln('annotation done')
# sys.exit(0)
## use. a_mpxml files
doer.sXmlFilenamePattern = doer.sLabeledXmlFilenamePattern
if options.iFoldInitNum or options.iFoldRunNum or options.bFoldFinish:
if options.iFoldInitNum:
"""
initialization of a cross-validation
"""
splitter, ts_trn, lFilename_trn = doer._nfold_Init(
lFold, options.iFoldInitNum, bStoreOnDisk=True)
elif options.iFoldRunNum:
"""
Run one fold
"""
oReport = doer._nfold_RunFoldFromDisk(options.iFoldRunNum,
options.warm)
traceln(oReport)
elif options.bFoldFinish:
tstReport = doer._nfold_Finish()
traceln(tstReport)
else:
assert False, "Internal error"
#no more processing!!
exit(0)
#-------------------
if lFold:
loTstRpt = doer.nfold_Eval(lFold, 3, .25, None)
import graph.GraphModel
sReportPickleFilename = os.path.join(sModelDir,
sModelName + "__report.txt")
traceln("Results are in %s" % sReportPickleFilename)
graph.GraphModel.GraphModel.gzip_cPickle_dump(sReportPickleFilename,
loTstRpt)
elif lTrn:
doer.train_save_test(lTrn, lTst, options.warm)
try:
traceln("Baseline best estimator: %s" %
doer.bsln_mdl.best_params_) #for GridSearch
except:
pass
traceln(" --- CRF Model ---")
traceln(doer.getModel().getModelInfo())
elif lTst:
doer.load()
tstReport = doer.test(lTst)
traceln(tstReport)
if lRun:
doer.load()
lsOutputFilename = doer.predict(lRun, sDocId)
traceln("Done, see in:\n %s" % lsOutputFilename)
def main(sModelDir, sModelName, options):
if options.use_ecn:
if options.ecn_json_config is not None and options.ecn_json_config is not []:
f = open(options.ecn_json_config[0])
djson = json.loads(f.read())
if "ecn_learner_config" in djson:
dLearnerConfig = djson["ecn_learner_config"]
f.close()
doer = DU_ABPTable_ECN(sModelName,
sModelDir,
dLearnerConfigArg=dLearnerConfig)
elif "ecn_ensemble" in djson:
dLearnerConfig = djson
f.close()
doer = DU_ABPTable_ECN(sModelName,
sModelDir,
dLearnerConfigArg=dLearnerConfig)
else:
doer = DU_ABPTable_ECN(sModelName, sModelDir)
elif options.use_gat:
if options.gat_json_config is not None and options.gat_json_config is not []:
f = open(options.gat_json_config[0])
djson = json.loads(f.read())
dLearnerConfig = djson["gat_learner_config"]
f.close()
doer = DU_ABPTable_GAT(sModelName,
sModelDir,
dLearnerConfigArg=dLearnerConfig)
else:
doer = DU_ABPTable_GAT(sModelName, sModelDir)
else:
doer = DU_ABPTable(sModelName,
sModelDir,
C=options.crf_C,
tol=options.crf_tol,
njobs=options.crf_njobs,
max_iter=options.crf_max_iter,
inference_cache=options.crf_inference_cache)
if options.rm:
doer.rm()
return
lTrn, lTst, lRun, lFold = [
_checkFindColDir(lsDir)
for lsDir in [options.lTrn, options.lTst, options.lRun, options.lFold]
]
traceln("- classes: ", doer.getGraphClass().getLabelNameList())
## use. a_mpxml files
doer.sXmlFilenamePattern = doer.sLabeledXmlFilenamePattern
if options.iFoldInitNum or options.iFoldRunNum or options.bFoldFinish:
if options.iFoldInitNum:
"""
initialization of a cross-validation
"""
splitter, ts_trn, lFilename_trn = doer._nfold_Init(
lFold,
options.iFoldInitNum,
test_size=0.25,
random_state=None,
bStoreOnDisk=True)
elif options.iFoldRunNum:
"""
Run one fold
"""
oReport = doer._nfold_RunFoldFromDisk(options.iFoldRunNum,
options.warm, options.pkl)
traceln(oReport)
elif options.bFoldFinish:
tstReport = doer._nfold_Finish()
traceln(tstReport)
else:
assert False, "Internal error"
#no more processing!!
exit(0)
#-------------------
if lFold:
loTstRpt = doer.nfold_Eval(lFold, 3, .25, None, options.pkl)
import crf.Model
sReportPickleFilename = os.path.join(sModelDir,
sModelName + "__report.txt")
traceln("Results are in %s" % sReportPickleFilename)
crf.Model.Model.gzip_cPickle_dump(sReportPickleFilename, loTstRpt)
elif lTrn:
doer.train_save_test(lTrn, lTst, options.warm, options.pkl)
try:
traceln("Baseline best estimator: %s" %
doer.bsln_mdl.best_params_) #for GridSearch
except:
pass
traceln(" --- CRF Model ---")
traceln(doer.getModel().getModelInfo())
elif lTst:
doer.load()
tstReport = doer.test(lTst)
traceln(tstReport)
if options.bDetailedReport:
traceln(tstReport.getDetailledReport())
import crf.Model
for test in lTst:
sReportPickleFilename = os.path.join(
'..', test, sModelName + "__report.pkl")
traceln('Report dumped into %s' % sReportPickleFilename)
crf.Model.Model.gzip_cPickle_dump(sReportPickleFilename,
tstReport)
if lRun:
if options.storeX or options.applyY:
try:
doer.load()
except:
pass #we only need the transformer
lsOutputFilename = doer.runForExternalMLMethod(
lRun, options.storeX, options.applyY, options.bRevertEdges)
else:
doer.load()
lsOutputFilename = doer.predict(lRun)
traceln("Done, see in:\n %s" % lsOutputFilename)
def main(DU_BAR):
version = "v.01"
usage, description, parser = DU_CRF_Task.getBasicTrnTstRunOptionParser(
sys.argv[0], version)
parser.add_option("--docid",
dest='docid',
action="store",
default=None,
help="only process docid")
# ---
#parse the command line
(options, args) = parser.parse_args()
# ---
try:
sModelDir, sModelName = args
except Exception as e:
traceln("Specify a model folder and a model name!")
_exit(usage, 1, e)
doer = DU_BAR(sModelName,
sModelDir,
C=options.crf_C,
tol=options.crf_tol,
njobs=options.crf_njobs,
max_iter=options.max_iter,
inference_cache=options.crf_inference_cache)
if options.docid:
sDocId = options.docid
else:
sDocId = None
if options.rm:
doer.rm()
sys.exit(0)
lTrn, lTst, lRun, lFold = [
_checkFindColDir(lsDir)
for lsDir in [options.lTrn, options.lTst, options.lRun, options.lFold]
]
# if options.bAnnotate:
# doer.annotateDocument(lTrn)
# traceln('annotation done')
# sys.exit(0)
## use. a_mpxml files
doer.sXmlFilenamePattern = doer.sLabeledXmlFilenamePattern
if options.iFoldInitNum or options.iFoldRunNum or options.bFoldFinish:
if options.iFoldInitNum:
"""
initialization of a cross-validation
"""
splitter, ts_trn, lFilename_trn = doer._nfold_Init(
lFold, options.iFoldInitNum, bStoreOnDisk=True)
elif options.iFoldRunNum:
"""
Run one fold
"""
oReport = doer._nfold_RunFoldFromDisk(options.iFoldRunNum,
options.warm)
traceln(oReport)
elif options.bFoldFinish:
tstReport = doer._nfold_Finish()
traceln(tstReport)
else:
assert False, "Internal error"
#no more processing!!
exit(0)
#-------------------
if lFold:
loTstRpt = doer.nfold_Eval(lFold, 3, .25, None, options.pkl)
import graph.GraphModel
sReportPickleFilename = os.path.join(sModelDir,
sModelName + "__report.txt")
traceln("Results are in %s" % sReportPickleFilename)
graph.GraphModel.GraphModel.gzip_cPickle_dump(sReportPickleFilename,
loTstRpt)
elif lTrn:
doer.train_save_test(lTrn, lTst, options.warm, options.pkl)
try:
traceln("Baseline best estimator: %s" %
doer.bsln_mdl.best_params_) #for GridSearch
except:
pass
traceln(" --- CRF Model ---")
traceln(doer.getModel().getModelInfo())
elif lTst:
doer.load()
tstReport = doer.test(lTst)
traceln(tstReport)
if lRun:
if options.storeX or options.applyY:
try:
doer.load()
except:
pass #we only need the transformer
lsOutputFilename = doer.runForExternalMLMethod(
lRun, options.storeX, options.applyY)
else:
doer.load()
lsOutputFilename = doer.predict(lRun)
traceln("Done, see in:\n %s" % lsOutputFilename)
def standardDo(self, options):
"""
do whatever is reuested by an option from the parsed command line
return None
"""
if options.rm:
self.rm()
return
lTrn, lTst, lRun, lFold = [_checkFindColDir(lsDir) for lsDir in [options.lTrn, options.lTst, options.lRun, options.lFold]]
# Validation set if any
try:
ratio_train_val = float(options.lVld[0])
lVld = []
if ratio_train_val <= 0 or 1.0 <= ratio_train_val: raise Exception("Bad ratio, not in ]0, 1[")
except:
ratio_train_val = None
lVld = _checkFindColDir(options.lVld)
#traceln("- classes: ", doer.getGraphClass().getLabelNameList())
## use. a_mpxml files
#doer.sXmlFilenamePattern = doer.sLabeledXmlFilenamePattern
if options.iFoldInitNum or options.iFoldRunNum or options.bFoldFinish:
if options.iFoldInitNum:
"""
initialization of a cross-validation
"""
splitter, ts_trn, lFilename_trn = self._nfold_Init(lFold, options.iFoldInitNum, test_size=0.25, random_state=None, bStoreOnDisk=True)
elif options.iFoldRunNum:
"""
Run one fold
"""
oReport = self._nfold_RunFoldFromDisk(options.iFoldRunNum, options.warm, options.bPkl)
traceln(oReport)
elif options.bFoldFinish:
tstReport = self._nfold_Finish()
traceln(tstReport)
else:
assert False, "Internal error"
return
if lFold:
loTstRpt = self.nfold_Eval(lFold, 3, .25, None, options.bPkl)
sReportPickleFilename = os.path.join(self.sModelDir, self.sModelName + "__report.txt")
traceln("Results are in %s"%sReportPickleFilename)
GraphModel.gzip_cPickle_dump(sReportPickleFilename, loTstRpt)
elif lTrn or lTst or lRun:
if lTrn:
tstReport = self.train_save_test(lTrn, lTst, lVld, options.warm, options.bPkl
, ratio_train_val=ratio_train_val)
try: traceln("Baseline best estimator: %s"%self.bsln_mdl.best_params_) #for GridSearch
except: pass
traceln(self.getModel().getModelInfo())
if lTst:
traceln(tstReport)
if options.bDetailedReport:
traceln(tstReport.getDetailledReport())
elif lTst:
self.load()
tstReport = self.test(lTst)
traceln(tstReport)
if options.bDetailedReport:
traceln(tstReport.getDetailledReport())
for test in lTst:
sReportPickleFilename = os.path.join('..',test, self.sModelName + "__report.pkl")
traceln('Report dumped into %s'%sReportPickleFilename)
GraphModel.gzip_cPickle_dump(sReportPickleFilename, tstReport)
if lRun:
# if options.storeX or options.applyY:
# try: self.load()
# except: pass #we only need the transformer
# lsOutputFilename = self.runForExternalMLMethod(lRun, options.storeX, options.applyY, options.bRevertEdges)
# else:
self.load()
lsOutputFilename = self.predict(lRun, bGraph=options.bGraph)
traceln("Done, see in:\n %s"%lsOutputFilename)
else:
traceln("No action specified in command line. Doing nothing... :)")
return
