STAR HardProbes PWG

[David Stewart <0ds.johnny AT gmail.com>]

Hi Rongrong,

Yes, you have summarized nicely. In taking the efficiency, I just take the ratio of the pT-integrated pre-corrected data to the pT-integrated corrected data; that ratio depends only on the shape of the pre-corrected data. If there were twice as many tracks pre-correction with the same shape, the resulting "efficiency" would be the same. 

On Fri, Jan 20, 2023 at 2:00 PM Ma, Rongrong <marr AT bnl.gov> wrote:

Hello David

Thanks for your response. 

I am not sure about "The result is that when correcting data-at-ZDCx-bin-A with embedding-at-ZDCx-bin-B, the only thing that matters is ZDCx-bin-B and not ZDCx-bin-A.". Since the amount of pileup is different in different data-at-ZDCx-bin-A bins, the corrected spectrum (including both shape and yield) must also depend on data-at-ZDCx-bin-A, right? Or you are only talking about shape here?

Let me try one more time to see if I understand. For each bin of (embedding-at-ZDCx-bin-B, data-at-ZDCx-bin-A), the efficiency is calculated using the spectrum shape from data-at-ZDCx-bin-A and applying it to the embedding events in embedding-at-ZDCx-bin-B. Is this correct? Given that the pT integrated efficiency is the same for different data-at-ZDCx-bin-A bins, it means the pT spectrum shape is the same for these bins which indicates the UP and primary tracks have similar pT shapes. 

Best

Rongrong

On Jan 20, 2023, at 1:39 PM, David Stewart <0ds.johnny AT gmail.com> wrote:

--sorry final clarification: "in each data-ZDCx bin using the embedding with the same corresponding embedding-zdcx bin (i.e. the same ZDCx range)" 

On Fri, Jan 20, 2023 at 1:36 PM David Stewart <0ds.johnny AT gmail.com> wrote:

Hi Rongrong,

I've got a longer response in draft to Nihar's questions. However, quickly here:

I took the pT efficiency responses from the embedding at different ZDCx bins in the embedding. Those efficiencies drop with increasing ZDCx. Then I used those efficiencies, and corrected the pT spectra of data in different ZDCx bins. The result is that when correcting data-at-ZDCx-bin-A with embedding-at-ZDCx-bin-B, the only thing that matters is ZDCx-bin-B and not ZDCx-bin-A. The figure that I showed was simply the efficiency calculated in this way using ZDCx-bin-A as the Y-axis and ZDCx-bin-B as the X axis. As Joern pointed out, this also a complicated way to indicate that the pT spectra of pileup (of which there is more at high ZDCx-bin-A) is about the same as the actual primary tracks. This is useful to us to know, because when correcting the tracks for detector efficiency, it doesn't matter how much PU is present, because the PU tracks and primary tracks will both correct with the same efficiency. Therefore algorithmically, we correct all detector level tracks (primary+PU together) in each ZDCx bin using the embedding for the same ZDCx, and then we look how the total number of tracks varies with ZDCx to determine how much must be PU.

Hope this helps,

Dave

On Fri, Jan 20, 2023 at 10:29 AM Ma, Rongrong via Star-hp-l <star-hp-l AT lists.bnl.gov> wrote:

Hello Veronica, Dave

I would like to follow up on this topic as I still do not fully understand y-axis either. 

What do you mean by "in doing the embedding, we do get a different ZDCx"? I would see that the ZDCx distributions are different between data and embedding, but for a given event there is only one ZDCx as recorded by ZDC during data taking, right? So maybe the question is how exactly do you obtain ZDCx_embed and ZDCx_data for a given event that goes into this efficiency plot. Could you show a snippet of  your code?

Thanks. 

Best

Rongrong

On Jan 20, 2023, at 1:18 AM, Veronica Verkest via Star-hp-l <star-hp-l AT lists.bnl.gov> wrote:

1) SLide#6:
_ What is Y-axis "ZDCX-bin data"? how do you calculate efficiency value
from data? (Not clear)

_ are these efficiency values are pT- and eta- integrated one?

These efficiency values are integrated over pT and eta. This plot might be more visually intuitive if you ignore the 7th bin on the x-axis, as this is just the average of the bins to the left. There is a typical pT-dependence shape in the curves and some slight dependence on eta (the asymmetry of the UE means different populations and therefore different efficiency for different eta-ranges in the TPC). The data-ZDCx bins were clear to me (first seeing the plot) as it is directly from the detector, but I had to reconcile to myself that, in doing the embedding, we do get a different ZDCx. The significance here is that the efficiency curves do not vary significantly as a function of the embedding ZDCx. This shows us that all tracks (real and PU) can be efficiency-corrected the same way, as the embedding ZDCx does not discriminate.

_______________________________________________
Star-hp-l mailing list
Star-hp-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-hp-l

--

David Stewart

Postdoctoral Fellow | Department of Physics, Wayne State University

--

David Stewart

Postdoctoral Fellow | Department of Physics, Wayne State University

--

David Stewart

Postdoctoral Fellow | Department of Physics, Wayne State University