Hello,<br><br>I also observed results similar to the one described by Hyupwoo, I do not remember exact number but it was big too.<br><br>I used the same analysis than Will to interpret this result, however here you take into account only Q2. You have to double the effect because the binning is two dimensional and similar effects is expected in W (or whatever other variable used). Moreover the binning is thinner where most of the events are concentrated and that enhance the expected effect.<br>
<br>My conclusions, seeing such big bin migration, were that it was under control as long as it remained in the direct neighbouring bin (I checked this). But it is also a clear indication that the binning is small enough (or maybe even already too small) and there is no use to reduce it more.<br>
<br>Raphael<br><br><div class="gmail_quote">On Wed, Aug 24, 2011 at 9:48 PM, William Brooks <<a href="mailto:william.brooks@usm.cl">william.brooks@usm.cl</a>> wrote:<br><blockquote class="gmail_quote">
Dear Hyupwoo et al,<br>
<br>
These are very interesting results, and surprising.<br>
<br>
If we can be more quantitative about the bin size, what bins did you use? in the plots such as <a href="http://www.pas.rochester.edu/%7Ehwlee/eg2/el-acc/C_Q2.html">http://www.pas.rochester.edu/~hwlee/eg2/el-acc/C_Q2.html</a>, it looks like 50 MeV per bin. I think the relevant way to think about this is bin width divided by resolution. I note that, for example, for charged particles the resolution is very good, but for charged particles that decay (like pions), the resolution has long tails that are larger at low momentum. Typical CLAS missing momentum "resolution" is 20 MeV. Typical decay fractions at high energies like 5 GeV beam might be 10%, e.g., 10% of the pions decay somewhere within the drift chamber volume, distorting the vertex, momentum, and angle resolution. So I expect the resolution is some kind of Gaussian-like distribution of RMS 20 MeV, with very long tails whose integral is 10% of the total area. For 50 MeV bins, this will not produce a 60% purity, it should be closer to 80-90%. So it seems that you have discovered something in!<br>
teresting and surprising which should be investigated further.<br>
<br>
Thanks,<br>
<br>
- Will<br>
<div><div></div><div class="h5"><br>
<br>
<br>
<br>
On Aug 24, 2011, at 3:19 PM, <a href="mailto:hwlee@pas.rochester.edu">hwlee@pas.rochester.edu</a> wrote:<br>
<br>
> Not so much for today for me...<br>
><br>
> I've checked the purities for each bin for acceptance calculations in<br>
> order to figure out how big the migrations are.<br>
> Quite a big amount of bins are having <50% purity and the average of<br>
> purities is ~60%...<br>
> This doesn't improve a lot with coarser binning.<br>
> I am checking what's the main reason for this big migrations.<br>
><br>
> I don't know if there were something more to discuss but for me that's all<br>
> for today.<br>
><br>
> Best,<br>
> Hyupwoo<br>
><br>
>> I am available if Hyupwoo is interested in discussing something. He is<br>
>> joining from home today if we have a meeting. -Steve<br>
>><br>
>>> Dear EG2,<br>
>>><br>
>>> This is a busy week at JLab, with the PAC meeting as well as other<br>
>>> possibly conflicting meetings. Do people want to have an EG2 meeting? I<br>
>>> am available, but not sure who else is interested.<br>
>>><br>
>>> Thanks,<br>
>>><br>
>>> - Will<br>
>>><br>
>>><br>
>>> _______________________________________________<br>
>>> Eg2 mailing list<br>
>>> <a href="mailto:Eg2@jlab.org">Eg2@jlab.org</a><br>
>>> <a href="https://mailman.jlab.org/mailman/listinfo/eg2">https://mailman.jlab.org/mailman/listinfo/eg2</a><br>
>>><br>
>>><br>
>><br>
>><br>
>><br>
>> _______________________________________________<br>
>> Eg2 mailing list<br>
>> <a href="mailto:Eg2@jlab.org">Eg2@jlab.org</a><br>
>> <a href="https://mailman.jlab.org/mailman/listinfo/eg2">https://mailman.jlab.org/mailman/listinfo/eg2</a><br>
>><br>
>><br>
><br>
><br>
><br>
> _______________________________________________<br>
> Eg2 mailing list<br>
> <a href="mailto:Eg2@jlab.org">Eg2@jlab.org</a><br>
> <a href="https://mailman.jlab.org/mailman/listinfo/eg2">https://mailman.jlab.org/mailman/listinfo/eg2</a><br>
<br>
<br>
_______________________________________________<br>
Eg2 mailing list<br>
<a href="mailto:Eg2@jlab.org">Eg2@jlab.org</a><br>
<a href="https://mailman.jlab.org/mailman/listinfo/eg2">https://mailman.jlab.org/mailman/listinfo/eg2</a><br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br>Raphaël Dupré<br><br><a href="http://www.jlab.org/%7Edupre/">Web Page</a><br><br>Argonne National Laboratory<br><br>Université Claude Bernard Lyon 1<br>