I took Brads advice and added my TDC cuts one by one to the BB.cer.asum_c variable (BB cerenkov adc sum) and seen how many events were left after applying cuts. For the TDC cuts I used DBB.BBcerT%02d[0]<400.0, where %02d is the TDC corresponding to the mirror that I want so %02d runs from 1 to 20 and chan 400 was selected so that I could quickly make the cuts and not worry about seperate TDC ranges all in time peaks are near chan 200, so timing peak is included in chan 400 cut. Here is a list of the results, where the arrow (T1->T3 say) means T1 | | T2 | | T3 and T1,T5 meand T1 | | T5 and no arrow or comma is just the individual TDC cut: <br>
<br>'OR' ed TDCs (DBB.BBcerT) evnets<br><br>No TDC cut 21376<br>T1 5208<br>T1,T2 6971<br>
T1->T3 8298<br>T1->T4 8672<br>T1->T5 7968<br>T1->T6 5001<br>
T1->T7 4577<br>T1->T8 3871<br>T1->T9 3333<br>T1->T10 2730<br>
T1->T11 584 <br>T1->T12 35<br>T11 1014<br>T11->T12 107<br>T12 323<br>
T13 910<br>T11,T13 284<br>T11->T14 51<br>T15 1344<br>T14,T15 367<br>
<br><br>So it seems that there is a big event fall off when the RHRS side of the TDCs is applied. I know that there is more background on the beamline side because of the higher rates, but I don't think the events should fall that much. <br>
<br>I Also made TDC cuts on BB.cer.asum_c using the BB.cer.t[ ] TDC variables from the THaCherenkov class, here are the results:<br><br>TDC Cut events<br><br>none 21376<br>T1 11810<br>
T1->T2 16273<br>T1->T3 18782<br>T1->T4 20022<br>T1->T5 20688<br>T1->T6 21102<br>T1->T7 21243<br>T1->T8 21304<br>
T1->T9 21339<br>T1->T10 21355<br>T1->T11 21371<br>T1->T12 21376<br>T1->T20 21376<br><br>This is what I was expecting. I am not too sure what the problem is. I think that I am going to form the DBB TDC cut by looping over Ndata hits as Brad commented below.<br>
<br><br>-Matt<br><div class="gmail_quote">On Sun, Jan 31, 2010 at 11:22 AM, Brad Sawatzky <<a href="mailto:brads@jlab.org">brads@jlab.org</a>> wrote:<br><blockquote class="gmail_quote">
<div class="im">On Fri, 29 Jan 2010, MATTHEW R POSIK wrote:<br>
<br>
> Ok, so I just tried using the cut<br>
><br>
> DBB.BBcerT11[0]<400.0||DBB.BBcerT12[0]<400.0||||DBB.BBcerT13[0]<400.0||DBB.BBcerT14[0]<400.0||DBB.BBcerT15[0]<400.0<br>
><br>
> and applied it to BB.cer.asum_c[0]<br>
<br>
</div>Only index array variables. BB.cer.asum_c is a scalar. It would be far<br>
better if ROOT just threw an error and crashed on bugs like that rather<br>
than attempting to guess "what you mean". If anyone can find an option<br>
that would make ROOT behave that way, PLEASE tell me! Usually ROOT<br>
guesses correctly, but it encourages sloppy code practice. When errors<br>
do crop up they're hard to debug...<br>
<br>
Bottom line is this:<br>
- always use an explicit index on array variables<br>
- never index a scalar variable<br>
- if you don't know if a variable is an array or scaler, then find out<br>
using T->Print(<var_name>), or equivalent<br>
<div class="im"><br>
> The result was out of ~500000 only 18 were left. I checked all the<br>
> TDCs and all have there in time peaks around 200-300, so having just<br>
> 18 events left doesn't make sense.<br>
<br>
</div>That is very strange. I recommend copying your generated cut string<br>
into a text file. Then, _by_hand_ at the root prompt, call<br>
T->Draw("..", ...) and add pieces of your cut step by step until you<br>
discover what part is throwing away your data.<br>
<br>
-- Brad<br>
<font color="#888888"><br>
--<br>
</font><div><div></div><div class="h5">Brad Sawatzky, PhD <<a href="mailto:brads@jlab.org">brads@jlab.org</a>> -<>- Jefferson Lab / Hall C / C111<br>
Ph: 757-269-5947 -<>- Pager: 757-584-5947 -<>- Fax: 757-269-7848<br>
The most exciting phrase to hear in science, the one that heralds new<br>
discoveries, is not "Eureka!" but "That's funny..." -- Isaac Asimov<br>
_______________________________________________<br>
d2n-analysis-talk mailing list<br>
<a href="mailto:d2n-analysis-talk@jlab.org">d2n-analysis-talk@jlab.org</a><br>
<a href="https://mailman.jlab.org/mailman/listinfo/d2n-analysis-talk">https://mailman.jlab.org/mailman/listinfo/d2n-analysis-talk</a><br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br>Matthew Posik<br>Email: <a href="mailto:posik@temple.edu">posik@temple.edu</a><br>Temple University Physics Dept.<br>Office: BA-319<br>Office #: 215-204-1331 <br>
<br>