[Clascomment] Electroproduction of the Lambda(1405)

[Reinhard Schumacher]

Hello Charles,

	Thank you for your comments on the paper that you posted on the CLAS 
review web site.  A revised version of the paper (v2) has been posted 
there.  Below we have replies to your questions.

Sincerely,
Reinhard and Haiyun
----------------------------------------------------------------------

Charles Hyde:
I have a few questions about the lineshape fits.

First of all, you say you tried both coherent and incoherent two-pole 
fits, with nearly equivalent results.  Did you try a coherent fit with
two nearly degenerate poles of strongly unequal widths?  This could 
produce a broad peak spanning the entire Lambda(1405) region with a 
narrower dip near the middle.  Depending upon the phase, the dip can be 
asymmetric. This might produce as good a fit as the incoherent fit, but 
with somewhat different conclusions.

 >>> Yes, we mention in the paper that a coherent fit was tried, just 
the way you describe it.  The result was that the fit was qualitatively 
(by eye) about the same quality as the incoherent fit, and the chi^2 was 
slightly, but not a lot, worse.  With the poor statistics we have, there 
is no discriminatory power to start teasing apart the effects of 
interferences.

Second,  In Figure 7 and 8, did you constrain the 8 fits (incoherent 
two-pole)to have the same pole positions and widths (even if they were 
globally free). From the figures, it appears you did not.  If not, I 
don't think it is a consistent fit and I don't think the chisquare is 
particularly significant, since you are (possibly) changing the fit 
parameters with every change in the bin size.

 >>> The revised paper uses fixed width and positions for each of the 
two BW line shapes across all the panels in the fit.  See Fig. 8.

Another comment on Fig 7.
The binning of the red fit curve is very strange. At large mass, it 
looks like a histogram with the same bin size as the data.
At low mass, it looks like a smooth curve with jagged breaks at each bin 
edge. This is very strange. What is going on?

There was a strange mistake due to the way RooFit plots results of fits. 
  That accounts for the wrong-looking scaling in some panels.  It has 
been fixed.  The smooth/jagged appearance is caused by the fact that the 
fit combines histogrammed information and smooth functional forms in the 
same fit function.  That's why the sometimes smooth-looking curve takes 
odd jumps.

-----------------------------------------------------------------------