Plotting Bugs: general

Bugs

1. When plotting scatter data (i.e., multiple y-values per x-value), x-errorbars are not plotted, even when the user requests x-errorbars.

2. One cannot plot convolved model components, only the overall convolved model stack.

3. Displayed x-errorbars are only correct if data are evenly binned.

Bugs

1. When plotting scatter data (i.e., multiple y-values per x-value), x-errorbars are not plotted, even when the user requests x-errorbars.

2. One cannot plot convolved model components, only the overall convolved model stack.

   sherpa> source = gauss[g]+pow[p]
   sherpa> lp source      # fine (unconvolved stack, g+p)
   sherpa> lp model       # fine (convolved stack, g+p)
   sherpa> lp g           # fine (_unconvolved_ component p)
   

   Workaround:

   sherpa> source = g
   sherpa> lp model
   sherpa> source = p
   sherpa> lp model
   

   If one wishes to overplot convolved components with the convolved stack (XSPEC-style counts plotting), it can be done via a Sherpa/S-Lang module workaround:

   sherpa> source = g
   sherpa> cong = get_mcounts()   # array of g amplitudes, convolved 
   sherpa> source = p
   sherpa> conp = get_mcounts()   # array of p amplitudes, convolved 
   sherpa> source = p+g
   sherpa> cons = get_mcounts()   # array of p+g amplitudes, convolved 
   sherpa> conx = get_axes()
   sherpa> print(conx)
   axistype         =  Energy
   axisunits        =  keV
   lo               =  Float_Type[95]
   hi               =  Float_Type[95]
   mid              =  NULL
   sherpa> mid = (conx.lo+conx.hi)/2.0  # make array of bin midpoints
   sherpa> () = curve(mid,cons)   # plot in ChIPS
   sherpa> () = curve(mid,conp)
   sherpa> () = curve(mid,cong)
   

3. Displayed x-errorbars are only correct if data are evenly binned.