RHICf Sub-reconstruction tool (StRHICfPID)#519
RHICf Sub-reconstruction tool (StRHICfPID)#519plexoos merged 6 commits intostar-bnl:mainfrom ggfdsa10:RHICfPID
Conversation
1. Particle Identification reconstruction
| float mPlateE[kRHICfNtower][kRHICfNplate]; | ||
| float mPlateSumE[kRHICfNtower]; | ||
| float mL20[kRHICfNtower]; | ||
| float mL90[kRHICfNtower]; |
There was a problem hiding this comment.
This is not an IO class AFAICT. Why define these variables with single precision?
There was a problem hiding this comment.
Hi Dmitri,
Thank you for your comments.
This class is a reconstructor for PID and not an IO class.
RHICf saving classes, such as StRHICfHit and StRHICfPoint, include mostly single-precision (float) variables.
StRHICfPID also receives input data as mPlateE from the StRHICfHit class using the setPlateEnergy() function.
Therefore, I didn’t see the need to change it to double precision.
However, I think calculation functions such as calculateEquation() could be changed to double precision.
Because that function needs to be more precise.
If you think it would be better to change these variables to double precision, I will make the changes along with others.
Thank you
| if(!mPlateEIs){return kRHICfFatal;} | ||
|
|
||
| for(int it=0; it<kRHICfNtower; it++){ | ||
| for(int ip=0; ip<mPlateIdxNum; ip++){mPlateSumE[it] += layerSumEnergy(it, ip);} |
There was a problem hiding this comment.
See if you can use std::accumulate() here and in the next lines. It is likely to be more efficient than adding elements in a loop with if-statements...
There was a problem hiding this comment.
This line of code can be replaced with the following:
mPlateSumE[it] += (2*std::accumulate(mPlateE[it][0], mPlateE[it][9], 0));
mPlateSumE[it] += mPlateE[it][10];
mPlateSumE[it] += (2*mPlateE[it][11]);
mPlateSumE[it] += (4*std::accumulate(mPlateE[it][12], mPlateE[it][14], 0));
mPlateSumE[it] += (2*mPlateE[it][15]);
However, I’m not sure which version is better.
In my opinion, the original code lines are better than the replacement.
I think the original version seems more intuitive and simpler.
2. change the function to std::fill() in init() 3. add init() in constructor
|
All of the input data-checking functions for RHICf come from the StRHICfFunction class. Furthermore, all input data checkers simply ensure that the total number of input data is correct. For example, they check that the setPlateEnergy() function has been called as many times as there are plate energy arrays. For this reason, I don’t think these checkers need to be included in all of the RHICf tools, so I plan to remove them from all RHICf tools soon. |
|
I think this one is ready to merge. Dmitri, please go head. Thanks |
all of the checking function is removed these functions are NOT necessary to RHICf reconstruction. In more detail, see the comment below, (in PR #519 ) ----------------------------------------------------------------------------------------------------------------------- > All of the input data-checking functions for RHICf come from the StRHICfFunction class. However, when someone uses a merged RHICf tool, those functions or codes will always be frozen. > > Furthermore, all input data checkers simply ensure that the total number of input data is correct. For example, they check that the setPlateEnergy() function has been called as many times as there are plate energy arrays. > > For this reason, I don’t think these checkers need to be included in all of the RHICf tools, so I plan to remove them from all RHICf tools soon. Commit[d92b32d](d92b32d) was made for the same reason. -------------------------------------------------------------------------------------------------------------------------
3 participants
RHICf Particle Identification reconstruction tool
this used in StRHICfPointMaker