Log of /trunk/roofit/roostats/inc/RooStats/ToyMCSampler.h
Parent Directory
Revision
46083 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Thu Sep 20 13:57:37 2012 UTC (2 years, 4 months ago) by
moneta
File length: 12981 byte(s)
Diff to
previous 45032
Fix re-setting a new prior pdf for the nnuisance. Invalidate in this case the nuisanceparameter sampler
(Thanks to Cristoph Nageli)
Revision
45032 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Thu Jul 12 16:03:34 2012 UTC (2 years, 6 months ago) by
moneta
File length: 12813 byte(s)
Diff to
previous 44472
apply fixes for Windows.
Implementation of static function should be in *.cxx file
in stressRooStats fix sqrt (thanks to Bertrand)
Revision
44472 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Fri Jun 1 16:08:27 2012 UTC (2 years, 7 months ago) by
moneta
File length: 12841 byte(s)
Diff to
previous 44428
fix some Coverity issues (missing copy ctor or assignment, dead code, etc..)
from Gabriel: add error in significance in HypoTestResult
Revision
44428 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Fri Jun 1 08:08:39 2012 UTC (2 years, 7 months ago) by
moneta
File length: 12838 byte(s)
Diff to
previous 44366
apply patch provided by Tim Adye:
1) Save fit errors in detailed outputs' RooRealVars, rather than as separate
variables. If withErrorsAndPulls is specified, the asymmetric errors are also
saved. Work round bug #94908 by using RooTreeDataStore in
DetailedOutputAggregator and ToyMCImportanceSampler (can be removed when #94908
is fixed).
2) Fix detailed output dataset name and variable order (dsname_TS0 should be
first variable). Previously SamplingDistributions from importance sampling were
wrong (taken from the first global observable!).
3) Make detailed output for data available via
HypoTestResult::GetAllTestStatisticsData().
4) Don't save detailed output in HypoTestResult if it was not requested.
5) Rationalised the saving and restoring of parameters in ToyMCSampler. The POI
value in the workspace is not overwritten before calling the test stat (restores
behaviour from 5.32).
6) Remove unnecessary dataset copy in ToyMCImportanceSampler.
7) Fix ToyMCImportanceSampler double->int compiler warnings.
8) Fix minor memory leaks in HypoTestCalculatorGeneric, ToyMCSampler, and
ProfileLikelihoodTestStat.
9) Update streamer so PiecewiseInterpolation::setAllInterpCodes works for
objects created with ROOT 5.30.
Revision
44366 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Wed May 30 15:32:27 2012 UTC (2 years, 7 months ago) by
axel
File length: 13141 byte(s)
Diff to
previous 44157
Remove #include "Riostream.h" from roofitcore/inc/RooCacheManager.h, which pulls in using namespace std.
Deal with the consequences.
Revision
44157 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Mon May 7 16:41:44 2012 UTC (2 years, 8 months ago) by
moneta
File length: 13096 byte(s)
Diff to
previous 43199
merge changes from roostats development branch up to revision 44155
Modifications by Jakub Ademek and Max Baak :
ensure that ToyMCSampler can return TTree with detailed fit output.
See for a detailed description:
https://indico.cern.ch/getFile.py/access?contribId=0&resId=3&materialId=slides&confId=182211
Modifications by Sven Kreiss:
ProfileLikelihoodTestStat:
- added option to add errors and pulls to detailed output
Revision 44128 - Directory Listing
Modified Fri May 4 21:33:08 2012 UTC (2 days, 17 hours ago) by sven
ToyMCSampler, ToyMCStudy, SamplingDistribution, DetailedOutputAggregator:
- bugfixes for the naming of the result
- return toys with weight=1 when generating unweighted toys (not -1)
- make it optional to return errors and pulls with detailed output
- from Tim Adye: catch NaNs at the right place
ProfileLikelihoodTestStat:
- added option to add errors and pulls to detailed output
FrequentistCalculator (from Tim Adye):
- skip the profiling if there are no free nuisance parameters
--This line, and those below, wil
_M roostats
_M roostats/src
M roostats/src/HypoTestResult.cxx
M roostats/src/FrequentistCalculator.cxx
M roostats/src/HypoTestCalculatorGeneric.cxx
M roostats/src/RooStatsUtils.cxx
M roostats/src/ToyMCSampler.cxx
M roostats/src/ToyMCStudy.cxx
M roostats/src/ProfileLikelihoodTestStat.cxx
M roostats/src/SamplingDistribution.cxx
_M roostats/inc
M roostats/inc/LinkDef.h
M roostats/inc/RooStats/ProfileLikelihoodTestStat.h
_M roostats/inc/RooStats/ToyMCSamplerOld.h
M roostats/inc/RooStats/FrequentistCalculator.h
M roostats/inc/RooStats/HypoTestCalculatorGeneric.h
A + roostats/inc/RooStats/DetailedOutputAggregator.h
M roostats/inc/RooStats/RatioOfProfiledLikelihoodsTestStat.h
M roostats/inc/RooStats/ToyMCStudy.h
M roostats/inc/RooStats/TestStatistic.h
M roostats/inc/RooStats/SamplingDistribution.h
M roostats/inc/RooStats/HypoTestResult.h
A + roostats/inc/RooStats/MinNLLTestStat.h
M roostats/inc/RooStats/RooStatsUtils.h
M roostats/inc/RooStats/ToyMCSampler.h
Revision
43199 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Thu Mar 1 20:17:42 2012 UTC (2 years, 10 months ago) by
moneta
File length: 13022 byte(s)
Diff to
previous 42319
merge from the roostats development branch the new developments from Sven Kreiss:
new ToyMCImportanceSampler
new version of the ToyMCSampler that can use multiple test statistics and "detailed output" of those test statistics.
extension to the FrequentistCalculator that allows the user to manually specify the values for the signal nuisance parameters using SetConditionalMLEsNull/Alt().
make HypoTestResult managing now the SamplingDIstribution for null and alt
Revision
42256 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Fri Nov 25 19:39:46 2011 UTC (3 years, 1 month ago) by
moneta
File length: 12800 byte(s)
Diff to
previous 40959
from Sven Kreiss:
adapt ToyMCSampler to use the new autobinning generation from RooFit
add additional functions:
// name of the tag for individual components to be generated binned (=> see RooFit::GenBinned() option)
void SetGenerateBinnedTag( const char* binnedTag = "" ) { fGenerateBinnedTag = binnedTag; }
// set auto binned generation (=> see RooFit::AutoBinned() option)
void SetGenerateAutoBinned( Bool_t autoBinned = kTRUE ) { fGenerateAutoBinned = autoBinned; }
Increased ClassDef version to 3.
Default is the same as RooFit, ie autobinned is kTRUE as default and fGenerateBinnedTag = "".
Revision
40959 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Tue Sep 20 16:33:21 2011 UTC (3 years, 4 months ago) by
moneta
File length: 12171 byte(s)
Diff to
previous 40001
- add in MCMCCalculator SetChainParameters to define the parameters to be put in the chain
- fix a bug of computing p values in SamplingDistribution introduced in 40920
- ToyMCSampler - clear the cached GenContext and pdf lists when setting a new PDF
- add a unique name for teh SamplingDist objects created in the Inverter
- fix also the shading in the plots of SamplingDist objects and the plotting in the log scale
Revision
39975 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Sun Jun 26 22:49:46 2011 UTC (3 years, 6 months ago) by
moneta
File length: 12036 byte(s)
Diff to
previous 39945
merge from roostats dev branch :
- from Sven: new version of ToyMCSampler
- from Kyle: update tutorial StandardTestStatDistributionDemo
-from LM : new version of Inverter with rebuilding pValues distributions using toys
+ update comments
Revision
39945 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Fri Jun 24 14:06:56 2011 UTC (3 years, 7 months ago) by
wouter
File length: 10877 byte(s)
Diff to
previous 39928
o RooHistFunc,LinkDef3
- Apply same fix to streamer as done in RooHistPdf
to ajust proxyList
o RooAbsArg
- Add RooHistFunc as friend
o RooChi2Var
- Savannah bug fix #81582
o RooHistError
- Savannah bug fix #80991
o ToyMCSampler,ProfileLikelihoodTestStat
- Implement 'dual switch' for speed optimization: a switch per object
and a static switch for all objects (the 'or' is used). The default
remains off.
Revision
39817 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Sat Jun 18 12:49:41 2011 UTC (3 years, 7 months ago) by
wouter
File length: 9889 byte(s)
Diff to
previous 39391
o RooDataSet
- Introduce pool-based memory allocation scheme
- In ctor for composite datasets with Link() ensure that
order of observables is preserved to allow use of assignFast()
o RooRealSumPdf
- Use getNorm() in expectedEvents() rather than createIntegral()
o RooAbsCollection
- Do no allocate hash tables by default
- Introduce contents claim counter to counter 'foreign' ownership
claims of contents (needed for RooAbsOptTestStatistic::setData()
support)
- If collection is owning, only delete contents if claim count is also zero
- In snapshot() raise hash table threshold from 100 to 1000
o RooRefCountedList
- Disable hash tables by default
o RooStudyPackage
- Make study package own clone of incoming workspace
- Implement copy ctor
o RooAbsTestStatistic
- Implement 'master' setData() method. Forward to setDataSlave()
of self for monolithic likelihoods or forward data components
to likelihood components in case node is SimMaster. If data
is not fast-splittable, execute slow split.
o RooAbsData
- Implement 'live check' to be able to conclude if a given dataset
pointer points to a live or dead object
- Claim contents of _vars through new claim mechanism. Release
in dtor, allow _vars to delete contents if claim count is zero
- Add concept of owned component datasets to support fast-splitting
of dataset
- Add method canSplitFast() that indicate if fast split is available
(only possible if dataset was created as composite)
- Add method getSimData() that retrieves pointer to component
RooDataSet representation for a given index state
o RooObjCacheManager
- Add static mechanism that allows to deleted cached observables lists used for
cache optimization triggers to be deleted on a sterilize() action. This
is useful on toy studies with setData() operation, that would otherwise
collect very large lists of observable clones during the toy cycle.
o RooAbsPdf
- Set contraint term to ADirty to avoid synchronization problems when using
setData() in the likelihood
o RooAbsArg
- In clone operations set hash table threshold at 1000 nodes
- In getObservables() use pass-by-value iterators rather than heap iterators
to making this faster. Also make getObservables() non-virtual, as it is in
fact not overloaded in any subclass (anymore). Also remove looping over
branch nodes to call getObservablesHook() as no class implements method
anymore.
o RooAddition
- Implement a setData() method that forwards to its components. Allows transparent
operation of setData() on likelihood objects with constraint terms
o RooLinkedList
- Add method that returns iterator by value (simply forwards to similar implementation
in RooLinkedList)
o RooSimGenContext
- Move calculation of fraction table from ctor to dedicated updateFraction() method
- Before each generation sequence call updateFractions() so that multi-generation
sequence are robust against parameter changes between calls
- In createDataSet() always make a composite dataset if the index is present, as this
allows faster use in toymc operations
o RooAbsOptTestStatistic
- Initialize all variables in default ctor
- In dtor and ctor handle ownership of dataset observables, on which we put a claim
(i.e. we need to delete contents if the test statistics survives the dataset)
- Reimplement existing setData() as setDataSlave(), and add claim management on
dataset observables. When attached new data to function, request that object caches
clear full content in sterilize() operation to avoid buildup of cache objects in
long toy runs.
- In setData() adjust event count of test statistics, for correct functioning of extended
terms after update
o RooArgSet
- In ctor for list and a single object, only add object if not already in list
(is already checked, but new procedure preserves order of list elements in case
object was already present)
o LinkDef
- Add list<RooAbsData*> and map<string,RooAbsData*> to support streaming of composite datasets
o RooAbsStudy
- Implement clone method to support polymorphic array cloning
o RooAbsReal
- Add interface method setData()
o RooProfileLL
- When storing coordinates of absolute minimum, store only those that are non-constant.
Otherwise option to restart each minimization from known absmin may alter constant
parameters
Revision
33206 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Mon Apr 26 09:59:06 2010 UTC (4 years, 8 months ago) by
moneta
Original Path:
trunk/roofit/roostats/inc/ToyMCSampler.h
File length: 10815 byte(s)
Diff to
previous 31276
merge fixes applied in roostats branch since 5.26 to current revision
ProfileLikelihoodCalculator:
in GetHypoTest clone the null parameters to avoid requesting the user to do this
LikelihoodInterval:
-fix drawing of horizontal green line when interval is outside the range
- return a boolean flag in Upper/LowerLimit with status of limit search
HybridCalculator: fix for test statistic = 3 (profile likelihood)
LikelihoodIn
Revision
29158 -
(
view)
(
download)
(
as text)
(
annotate)
-
[select for diffs]
Modified
Tue Jun 23 07:59:38 2009 UTC (5 years, 7 months ago) by
brun
Original Path:
trunk/roofit/roostats/inc/ToyMCSampler.h
File length: 10900 byte(s)
Diff to
previous 29107
From Wouter:
Fix when running the test with ACLIC.
However: there remains a peculiar problem with CINT and RooGlobalFunc.h
The latter defines a number of functions in namespace RooFit
but any code that needs to use it in CINT, must be written as follows
-------- demo.C ----------
#ifndef __CINT__
#include "RooGlobalFunc.h"
#endif
using namespace RooFit ;
void demo() {
// Function defined in ns RooFit by RooGlobalFunc.h
LineColor(kRed) ;
}
-------------------------
If the #ifndef is removed, the code does _not_ work in CINT...
With the #ifndef it works and its contents is visible.
I do not understand why. It seems wrong.
RooGlobalFunc.h does not contain anything funny or unusual...
This form allows you to request diffs between any two revisions of this file.
For each of the two "sides" of the diff,
enter a numeric revision.
