LCG Physicist Interfaces project

1
LCG Physicist Interfaces project
240 Status and Plans of the LCG PI project
Vincenzo Innocente, Lorenzo Moneta, Andreas
Pfeiffer CERN, EP/SFT H.C.Lee, W.L.Ueng Academia
Sinica Computing Centre, Taiwan
http//cern.ch/pi
PI -- A project of the LCG Application Area
AIDA Using Standards for Data Analysis
AIDA Abstract Interfaces for Data Analysis (see
379) Open collaboration started in 1999 to
discuss and agree on a common set of
(Abstract) Interfaces Flexibility and
extensibility User decides at runtime which
implementation to use No change in code
required ! Interoperability Modules interact
through interfaces only, so mixing of
different implementations is easy, even across
languages using JAIDA/AIDAJNI. Other AIDA
related presentations 213, 311, 333, 349,
379, 399, 400, 427
Mandate of the PI project AIDA Review, adapt,
extend Interface to Data Analysis Root
implementation of AIDA Provide an
implementation of the Interfaces to Data
Analysis, as defined by AIDA, based on Root
(using a wrapper to Root) AIDA interface to SEAL
and POOL services Use SEAL and POOL to provide
services such as plug-in manager, object
whiteboard and persistency. Blueprint compliant
Analysis tool set Statistics analysis tools
based on AIDA interface Mainly external
contributions Maintenance issue Re-use dont
re-invent ! Integration of existing s/w
Integration of PI s/w into experiments
frameworks
The PI testing suite
The PI AIDA Proxy layer

• Histogram Sources
• Filling with fixed and random data
• Copying from different implementation
• Reading from XML/ROOT file

Hurng-Chun Lee Academia Sinica Computing Centre,
Taiwan
The PI AIDA-Proxy layer All AIDA functionality
is available Excluding ITree Easy to use
Hide factories from users Value semantics
Implemented operator and Copy between
implementations Copying AIDA-native to
AIDA-Root and vice versa Choose implementation at
runtime User decides implementation when
constructing the objet Easier integration with
other frameworks Objects are managed by the
user (not by AIDA Tree)
Purpose Check the functionalities of AIDA
Proxy Check the consistencies between different
histogram implementations through the AIDA
Proxy layer Develop a unit test framework for
AIDA Proxy on which test-cases can be easily
adapted 1164 CppUnit assertions only 104 (lt10)
failures Failures are due to implementation
differences Root takes the bin centers instead
of the values to calculate the global mean in
H3D and Profiles Root doesnt store the
binMean Error treatment in Profile histograms
is different
2
LCG Physicist Interfaces project
240 Status and Plans of the LCG PI project
Vincenzo Innocente, Lorenzo Moneta, Andreas
Pfeiffer CERN, EP/SFT H.C.Lee, W.L.Ueng Academia
Sinica Computing Centre, Taiwan
http//cern.ch/pi
Using Python as the integration bus simplifies
integration Integrating with HippoDraw
Using PyRoot from the SEAL project (see also
175, 222) Integrating with Root
from AidaProxy import create 1D and 2D
histogram and plot them in a root canvas
using fixed bin histograms as root doesnt know
about clouds h1 Histogram1D("h1",LogNormal
Distribution",100,0,10) h2 Histogram2D("h2",Gau
ssian Distribution",100,0,100,100,0,100) fill
histograms with random gaussian points import
random for i in range(0,5000)
h1.fill(random.lognormvariate(0,1)) h2.fill(
random.gauss(45., 10.) , random.gauss(60., 20.) )
create root canvas from rootPlotter import
rpRootPlotter("AIDA to Root example",200,10,700,
800) set two canvas zones and options
rp.canvas.Divide(2,1) rp.canvas.cd(1)
rp.plot(h1) rp.canvas.cd(2) rp.plot(h2)
from AidaProxy import create 1D and 2D cloud
(unbinned histogram) and plot them using
hippoDraw c1 Cloud1D("MyCloud", "Log Normal
Distribution") c2 Cloud2D("MyCloud2",
"Gaussian Distribution") fill histograms with
random gaussian points import random for i in
range(0,5000) c1.fill( random.lognormvariate
(0,1) ) c2.fill( random.gauss(45., 10.) ,
random.gauss(60., 20.) ) create HippoPLotter
from hippoPlotter import plHDHippoPlotter()
plot cloud and create a binned histogram on a
region plHD.plot(c1,'Histogram','x','Entries/bin'
) h1 Histogram1D(h1,Log Normal
(Binned),100,0,5) c1fillHistogram(h1) plHD.plot(
h1) now plot 2D clouds plHD.plot(c2,'Scatter
Plot','x','y') plHD.plot(c2,'Color
Plot','x','y')
Cross-language interoperability Integrating with
JAIDA
from AidaProxy import create 1D and 2D cloud
and plot them in hippoDraw h1
Histogram1D("h1",LogNormal Distribution",100,0,10
) c2 Cloud2D("MyCloud2", "Gaussian
Distribution") fill histograms with random
gaussian points import random for i in
range(0,5000) h1.fill(random.lognormvariate(
0,1)) xval 1.random.gauss(45., 10.)
yval 1.random.gauss(60., 20.)
c2.fill(xval,yval,1.) create Plotter using
JAIDA pl pi_aida.Plotter() pl.createRegions(1,2)
pl.region(0).plot(h1) pl.show()
pl.region(1).plot(c2) pl.show()