On Mon, 16 Apr 2001, Christian Holm Christensen wrote:
> Bjorn, I've asked you this before but I can't remember getting a
> reply: What's the data members BrVertex::fVertex_sigma_<x|y> and
> BrVertex::fVariance<X|Y>? Sigma (or standard diviation or square root
> of the variance) is an intrinsic property of some probablity
> distribution, while "sample variance" (s^2) and "sample standard
> diviation" (s) is meassured quantaties. If, what you mean by these two
> members is that fVertex_sigma_<x|y> = sqrt(fVariance<X|Y>), then one
> of the sets is redundant (fVertex_sigma_<x|y>) and should not be
> present. A method (GetDiviationX() const { return
> TMath::Sqrt(fVarianceX); }) is what you need.
Ho Christian.
Those variables stem from the algorithm used in
BrTPMClusterVertexModule. If you recall, it makes a histogram of
projections of lines between clusters and then fits a gaussian to this
distribution (done individually for the y and z coords). fVertex_sigma_y
is the sigma of this fit in the y dir and similarily for z.
I believe I discussed the horrors of fVariance in a mail a while back -
its name is completely misleading. It should have been changed ages ago
along with a lot of stuff in the ClusterModule, but since the code works I
have concentrated on writing my thesis instead. (If Trine can get me a
PhD-position I might actually get to fix the code too ;-) Anyway -
fVariance is actually an estimate of the error on the mean of the
measurement(!!), currently calculated as sigma^2/n_entries where sigma is
from the fit and n_entries is from the histogram. This is of course also a
horrible calculation since it mixes a fit with a measurement and I have
actually started to change it on several occations but something has
always turned up...
My plan was to have the following data members in BrVertex from the TPC
vertex finders:
fCVtxFitSigma_y
fCVtxFitSigma_z
-- the sigmas from the fit. They are needed to cut away noisy events, at
least for this years data.
fCVtxError_y
fCVtxError_z
-- the estimated sample error on the mean, calc. by the textbook
estimator for a gaussian sigma. This is the main cvtx cut.
fTVtxChisq
-- reduced chisq from the 3D least--squares used by the
TrackVertexFinder. Main tvtx cut.
...in addition to obvious stuff like the position, fVertexFound etc. Not a
big change from now, but the names will at least be correct. On a slightly
longer timescale I intend to rework both of the algorithms since I am now
better at manipulating data with ROOT than I was 18 months ago... I don't
know if the efficiency and presicion can change much but the speed and
ease--of--use sure can.
Hope that was more or less clear.
As for the BB vertex, I also don't know where that shift may come
from... Vertices are Strange and Bewildering Beasts.
Ping :-)
------------------------------------------------
Bjorn H. Samset
Master-student in Heavy Ion physics
Mob: +47 92 05 19 98 Office: +47 22 85 77 62
Adr: Kri 2A709 Sognsveien 218 0864 Oslo
This archive was generated by hypermail 2b29 : Tue Apr 17 2001 - 10:10:45 EDT