ROOT version 6.02/05 has been released on February 9, 2015.
For more information, see:
The following people have contributed to this new version:
Bertrand Bellenot, CERN/SFT,
Dario Berzano, CERN/SFT,
Rene Brun, CERN/SFT,
Ioan Gabriel Bucur, CERN/SFT
Philippe Canal, FNAL,
Cristina Cristescu, CERN/SFT,
Olivier Couet, CERN/SFT,
Kyle Cranmer, NYU, RooStats,
Anders Eie, NTNU,
Gerri Ganis, CERN/SFT,
Andrei Gheata, CERN/Alice,
Wim Lavrijsen, LBNL, PyRoot,
Sergey Linev, GSI, http,
Anna-Pia Lohfink,
Pere Mato, CERN/SFT,
Lorenzo Moneta, CERN/SFT,
Axel Naumann, CERN/SFT,
Danilo Piparo, CERN/SFT,
Timur Pocheptsov, CERN/SFT,
Fons Rademakers, CERN/SFT,
Paul Russo, FNAL,
Joerg Stelzer, DESY/Atlas, TMVA,
Alja Tadel, UCSD/CMS, Eve,
Matevz Tadel, UCSD/CMS, Eve,
Eckhard von Toerne, University Bonn, ATLAS, TMVA,
Vassil Vassilev, CERN/SFT
Wouter Verkerke, NIKHEF/Atlas, RooFit,
Yue Shi Lai, MIT
The new interface TInterpreter::Declare(const char* code)
will declare the code to the interpreter with all interpreter extensions disabled, i.e. as “proper” C++ code. No autoloading or synamic lookup will be performed.
TIsAProxy’s constructor no longer take the optional and unused 2nd argument which was reserved for a ‘context’. This context was unused in TIsAProxy itself and was not accessible from derived classes.
Introduce GZIP compression on the server side. Now one can request JSON (or any other data) in zipped form like:
wget http://localhost:8080/Canvases/c1/root.json.gz
This solves problem with JSON using over network - such compressed file is about the same size as binary buffer.For that particular canvas (from hsimple.C example)
format | size |
---|---|
root.json |
12994 bytes |
root.json?compact=3 |
8695 bytes |
root.json.gz?compact=3 |
2071 bytes |
It is factor 4 less data, transmitted between server and client. “root.bin” request has also been modified. Now it is just data produced by TBufferFile without any additional headers.One can also compress such data with gzip:
wget http://localhost:8080/Canvases/c1/root.bin.gz
colz
in a command like nt->Draw("b:a:c>>h", "", "colz");
erased the histogram h
. (Jira report ROOT-4508).FUNC
for 2D histograms in the same way it is implmented for 1D. Ie: when the option FUNC
specified only the functions attached to the histogram are drawn.New drawing option VIOLIN
for 2D histograms from Davide Gerbaudo (davide.gerbaudo@gmail.com). A violin plot is a box plot that also encodes the pdf information at each point. Quartiles and mean are also represented at each point, with a marker and two lines.
In this implementation a TH2 is considered as a collection of TH1 along X (option VIOLIN or VIOLINX) or Y (option VIOLINY).
A solid fill style is recommended for this plot (as opposed to a hollow or hashed style).
Example:
{
Int_t nx(6), ny(40);
Double_t xmin(0.0), xmax(+6.0), ymin(0.0), ymax(+4.0);
TH2F* hviolin = new TH2F("hviolin", "Option VIOLIN example", nx, xmin, xmax, ny, ymin, ymax);
TF1 f1("f1", "gaus", +0,0 +4.0);
Double_t x,y;
for (Int_t iBin=1; iBin<hviolin->GetNbinsX(); ++iBin) {
Double_t xc = hviolin->GetXaxis()->GetBinCenter(iBin);
f1.SetParameters(1, 2.0+TMath::Sin(1.0+xc), 0.2+0.1*(xc-xmin)/xmax);
for(Int_t i=0; i<10000; ++i){
x = xc;
y = f1.GetRandom();
hviolin->Fill(x, y);
}
}
hviolin->SetFillColor(kGray);
hviolin->SetMarkerStyle(20);
hviolin->SetMarkerSize(0.5);
hviolin->Draw("VIOLIN");
c1->Update();
return c1;
}
#minus
sign is used instead, which improve the appearance of the plots and make them even better for publications.