Ostap decorates many ROOT.RooFit classes, adding more convinient methods to them.
All these classes have got set of additional python-like methods for iteration, extension, addition, elemtn access checking the content etc...
Also several methods to provide more coherent interfaces (e.g. add vs Add) are added.
These methods also have got the extended interface with many useful methods and operators, like
e.g. concatenation of datasets a+b and merging them a*c.
RooDataSet class also has go many methods, that are similar to those of ROOT.TTree, in particular project and draw:
dataset = ...
dataset.draw('mass','pt>1')
histo = ...
dataset.project ( histo , 'mass', 'pt>1' ) Many other methonds like statVar, sumVar , statCov , vminmax are also the same as for ROOT.TTree, see above.
s1 = dataset.statVar ('eff')
s2 = dataset.sumVar ('eff')
r = dataset.statCov ('eff','pt')
mn,mx = dataset.vminmax ('eff') The class RooFitResult get many decorations that allow to access fit results
result = ...
par1 = result.params() ## get all floating parameters
par2 = result.params( float_only = False ) ## all parameters
a,v = result.param ( 'a' ) ## par by name
a,v = result.param ( a ) ## par by RooFit object itself
p = result.a ## par as attribute
for par in result : print par ## iteration
for name,par in result.iteritems() : print par ## iteration
print result.cov ( 'a' , 'b' ) ## get the covariance submatrix
print result.corr ( 'a' , 'b' ) ## get the correlation coefficient Also the simple math with fiting parameters is supported
result = ...
s = result.sum ('S','B' ) ## S+B
d = result.divide ('S','B' ) ## S/B
s = result.subtract ('B','B1') ## B-B1
m = result.multiply ('A','B' ) ## A*B
f = result.fraction ('S','B' ) ## S/(S+B) Few simple operations are added to simplify the calculations with RooRealVar objects:
x = ROOT.RooRealVar( ... )
x + 10
x - 10
x * 10
x / 10
10 + x
10 - x
10 * x
10 / x
x += 2
x -= 2
x *= 2
x /= 2
x ** 3