Tadeusz Pytlos wrote:
> Rene,
> Thank you for your remarks. Maybe this thing is not worth
> to discuss, but in fact my problem is aesthetic nature.
> I calculated many TH2S("h2","h2",m,0,m)
> histos of integer and write them to file.
> Next I displayed them and saw that the positions pairs of integers
> are not integer!. I calculated TH2S("h2","h2",m+1,0,m) - the same.
> I thought that taking TH2F with right ranges
> could solve the problem, but it failed.
> Of course main reasons for these histos are next calculation,
> so I don't want to lost advantages of TH2.
> I didn't use TPolyMarker yet and I'm a little afraid whether
> it has the similar functionality.
> Is any reason that (my above examples) it's possible
> to draw point in TH1F as integer in the middle of class,
> but it's impossible to draw pairs of integer in TH2F in the same way?
> I'm not quite sure, but the same problem had been in PAW as well.
> Best wishes,
> Tadeusz
Tadeusz,
When a TH2 object is rendered as a scatter-plot (via
TH1::PaintScatterPlot) each cell is represented by a bunch of points
distributed randomly in the cell area. The number of points
is proportional to the cell content.
This is the best thing we can do because the original information
has been lost by definition.
You can use a TGraph, TPolyMarker or TNtuple/TTree objects
if you want to draw a number of points with their original precision.
Rene Brun
This archive was generated by hypermail 2b29 : Tue Jan 04 2000 - 00:34:38 MET