Hello all,
We received answers from the comments on: LHCB-PAPER-2014-012-001 - Measurement of the resonant and $CP$ components in B_s -> Jpsipipi
Here is the cds link of the paper: https://cds.cern.ch/record/1669578
And here the response from Sheldon: https://cds.cern.ch/record/1692365
I forward the original message.
Vasilis
-------- Original Message -------- Subject: LHCB-PAPER-2014-012-001-COMMENT-004: review received Date: Mon, 31 Mar 2014 23:28:30 +0200 From: CERN Document Server Submission Engine cds.support@cern.ch To: vasileios.syropoulos@cern.ch
Dear LHCb Colleague, Your comment LHCB-PAPER-2014-012-001-COMMENT-004 on draft LHCB-PAPER-2014-012-001 (entitled: 'Measurement of the resonant and $CP$ components in $\bar{B}^0 \to J/\psi\pi^+\pi^-$ decay') has been responded to by Sheldon Stone of the reviewers.
The review will soon be available at the following URL:
http://cds.cern.ch/record/1692365 Kind regards, CERN Document Server.
Hi Vasilis,
I have some follow-ups. See below. In particular, I am worried what they answer on l.164. They first extract the signal and background fractions from the B invariant mass, and then they constrain the B mass, but still use the same background and signal fractions. Does anyone know if there is a default chi2 cut in the mass-constrained fits?
Cheers Jeroen
We think you are referring to the fact that the fit best seen in Fig. 16 is somewhat above the curve at 980 MeV and slightly above around 924 MeV. We interpret these as statistical fluctuations and not indicative of any “oscillatory behavior”, since the width of the 980 is narrow enough that only about 25% of the peak number of events are present at 924 MeV (see LHCb-PAPER-2013-069 Interference of the f0(980) with other scalar states is allowed in the fit and taken into account. Adding NR improves the fit in that region, which gives quite large change on the ratio of f0(980)/f0(500). The limit is based on the NR fit, rather than the more restrictive fit without it. In addition, varying of f0(500) and f0(980) resonance parameters also changes the fit, and those are included in the syst. uncertainty. We also don’t agree that any further discussion here is warranted.
What is wrong with spending one or two sentences on this? We do not question the results. It would make the paper stronger when you say that this was carefully checked by varying the f0 resonant parameters, and that this pattern is compatible with a statistical fluctuation.
- l.5: “Cabibbo-suppressed”. Add the dash. - Not normal EB style
No. Please have a look in the EB guidelines.
- l.44: “by use of either IP requirements or detachment of the Jpsi from the primary vertex”. IP requirements of what? Of the Jpsi or its daughters? Please check. Also add meson after Jpsi. - We think the statement is clear as is and has been used previously. Adedd “meson”
If you think the statement is clear, you can answer our question. We guess the answer is that the IP requirement refers to the muons, and then the sentence should be revised.
- l.52: remove “four variables”. Every reader is able to add 1+3. - Prefer as is as we are introducing a list
Yes, but the Latin-numbered list is only three long. Please remove "four variables:".
- Eq.3+4: This should not be written in differential form. This makes it inconstant (and thus wrong) with Eqs. 5-9. In this form the amplitude A should be a differential itself since the full functional form it in there (except for the decay time). However, you do not write it as such in the other equations. - We disagree. Eq 5-9 compliment Eq. 3 &4 and are stated properly.
We still disagree with the current notation. Please consider revising.
- l.75: write “|q/p|” the same way as in the equations before. - This would be completely incorrect! We take the absolute value to be unity, not the phase.
We meant writing |q/p| instead of |p/q|.
- Eq.3+4: Since you make the assumptions for q/p and indirect CPV, please simplify these equations from the start. - It is important to show the reader what we are assuming from the general equations
That is not the point. You can still explicitly write the assumptions and give the simplified expressions.
- l.164: You constrain the B0 mass to the PDG value for the amplitude analysis. But in l.172 you use the signal fraction from the invariant B mass fit for the angular fit. This looks wrong. You cannot use that number after constraining to the B mass. Maybe you only constrain to the B mass for another fit, but this should be made clear in the text. - Our procedure is consistent. The invariant mass fit gives the signal and background fractions. We then can do whatever we wish to the events, constrain them etc… This does not change the signal and background fractions.
No, this seems to be really wrong then. As soon as you do a constrained fit on the B mass, you typically apply a chi2 cut, which changes the fractions. Even if you don’t, you need to check that not too many fits failed. And if you do all this correctly, please mention this in the text to avoid any confusion.
- l.296: “efficiency-corrected” add dash. - OK, also “background-subtracted” then
Indeed.
- Eq.34-37: These are all results, so remove the equation number. - OK for 34 & 37, the other two are not our results.
Does not matter if it is yours or not. The EB rule is that displayed results are not numbers.
- Eq.37: C.L -> CL (without dots). - Prefer as is
Check other papers. All of them have no dots.
Hey Jeroen,
I just submitted your comments. https://cds.cern.ch/record/1692901
Vasilis
On 04/02/2014 10:12 AM, Jeroen Van Tilburg wrote:
Hi Vasilis,
I have some follow-ups. See below. In particular, I am worried what they answer on l.164. They first extract the signal and background fractions from the B invariant mass, and then they constrain the B mass, but still use the same background and signal fractions. Does anyone know if there is a default chi2 cut in the mass-constrained fits?
Cheers Jeroen
We think you are referring to the fact that the fit best seen in Fig. 16 is somewhat above the curve at 980 MeV and slightly above around 924 MeV. We interpret these as statistical fluctuations and not indicative of any “oscillatory behavior”, since the width of the 980 is narrow enough that only about 25% of the peak number of events are present at 924 MeV (see LHCb-PAPER-2013-069 Interference of the f0(980) with other scalar states is allowed in the fit and taken into account. Adding NR improves the fit in that region, which gives quite large change on the ratio of f0(980)/f0(500). The limit is based on the NR fit, rather than the more restrictive fit without it. In addition, varying of f0(500) and f0(980) resonance parameters also changes the fit, and those are included in the syst. uncertainty. We also don’t agree that any further discussion here is warranted.
What is wrong with spending one or two sentences on this? We do not question the results. It would make the paper stronger when you say that this was carefully checked by varying the f0 resonant parameters, and that this pattern is compatible with a statistical fluctuation.
- l.5: “Cabibbo-suppressed”. Add the dash.
- Not normal EB style
No. Please have a look in the EB guidelines.
- l.44: “by use of either IP requirements or detachment of the Jpsi
from the primary vertex”. IP requirements of what? Of the Jpsi or its daughters? Please check. Also add meson after Jpsi.
- We think the statement is clear as is and has been used previously.
Adedd “meson”
If you think the statement is clear, you can answer our question. We guess the answer is that the IP requirement refers to the muons, and then the sentence should be revised.
- l.52: remove “four variables”. Every reader is able to add 1+3.
- Prefer as is as we are introducing a list
Yes, but the Latin-numbered list is only three long. Please remove "four variables:".
- Eq.3+4: This should not be written in differential form. This makes
it inconstant (and thus wrong) with Eqs. 5-9. In this form the amplitude A should be a differential itself since the full functional form it in there (except for the decay time). However, you do not write it as such in the other equations.
- We disagree. Eq 5-9 compliment Eq. 3 &4 and are stated properly.
We still disagree with the current notation. Please consider revising.
- l.75: write “|q/p|” the same way as in the equations before.
- This would be completely incorrect! We take the absolute value to be
unity, not the phase.
We meant writing |q/p| instead of |p/q|.
- Eq.3+4: Since you make the assumptions for q/p and indirect CPV,
please simplify these equations from the start.
- It is important to show the reader what we are assuming from the
general equations
That is not the point. You can still explicitly write the assumptions and give the simplified expressions.
- l.164: You constrain the B0 mass to the PDG value for the amplitude
analysis. But in l.172 you use the signal fraction from the invariant B mass fit for the angular fit. This looks wrong. You cannot use that number after constraining to the B mass. Maybe you only constrain to the B mass for another fit, but this should be made clear in the text.
- Our procedure is consistent. The invariant mass fit gives the signal
and background fractions. We then can do whatever we wish to the events, constrain them etc… This does not change the signal and background fractions.
No, this seems to be really wrong then. As soon as you do a constrained fit on the B mass, you typically apply a chi2 cut, which changes the fractions. Even if you don’t, you need to check that not too many fits failed. And if you do all this correctly, please mention this in the text to avoid any confusion.
- l.296: “efficiency-corrected” add dash.
- OK, also “background-subtracted” then
Indeed.
- Eq.34-37: These are all results, so remove the equation number.
- OK for 34 & 37, the other two are not our results.
Does not matter if it is yours or not. The EB rule is that displayed results are not numbers.
- Eq.37: C.L -> CL (without dots).
- Prefer as is
Check other papers. All of them have no dots.
Hello all,
We got a reply from Sheldon for the *second* round of comments. Here is the link http://cds.cern.ch/record/1692901 I forward the original message as well.
Vasilis
-------- Original Message -------- Subject: LHCB-PAPER-2014-012-001-COMMENT-010: review received Date: Wed, 2 Apr 2014 16:20:17 +0200 From: CERN Document Server Submission Engine cds.support@cern.ch To: vasileios.syropoulos@cern.ch
Dear LHCb Colleague, Your comment LHCB-PAPER-2014-012-001-COMMENT-010 on draft LHCB-PAPER-2014-012-001 (entitled: 'Measurement of the resonant and $CP$ components in $\bar{B}^0 \to J/\psi\pi^+\pi^-$ decay') has been responded to by Sheldon Stone of the reviewers.
The review will soon be available at the following URL:
http://cds.cern.ch/record/1692901 Kind regards, CERN Document Server.
On 04/02/2014 01:26 PM, V.Syropoulos wrote:
Hey Jeroen,
I just submitted your comments. https://cds.cern.ch/record/1692901
Vasilis
On 04/02/2014 10:12 AM, Jeroen Van Tilburg wrote:
Hi Vasilis,
I have some follow-ups. See below. In particular, I am worried what they answer on l.164. They first extract the signal and background fractions from the B invariant mass, and then they constrain the B mass, but still use the same background and signal fractions. Does anyone know if there is a default chi2 cut in the mass-constrained fits?
Cheers Jeroen
We think you are referring to the fact that the fit best seen in Fig. 16 is somewhat above the curve at 980 MeV and slightly above around 924 MeV. We interpret these as statistical fluctuations and not indicative of any "oscillatory behavior", since the width of the 980 is narrow enough that only about 25% of the peak number of events are present at 924 MeV (see LHCb-PAPER-2013-069 Interference of the f0(980) with other scalar states is allowed in the fit and taken into account. Adding NR improves the fit in that region, which gives quite large change on the ratio of f0(980)/f0(500). The limit is based on the NR fit, rather than the more restrictive fit without it. In addition, varying of f0(500) and f0(980) resonance parameters also changes the fit, and those are included in the syst. uncertainty. We also don't agree that any further discussion here is warranted.
What is wrong with spending one or two sentences on this? We do not question the results. It would make the paper stronger when you say that this was carefully checked by varying the f0 resonant parameters, and that this pattern is compatible with a statistical fluctuation.
- l.5: "Cabibbo-suppressed". Add the dash.
- Not normal EB style
No. Please have a look in the EB guidelines.
- l.44: "by use of either IP requirements or detachment of the Jpsi
from the primary vertex". IP requirements of what? Of the Jpsi or its daughters? Please check. Also add meson after Jpsi.
- We think the statement is clear as is and has been used previously.
Adedd "meson"
If you think the statement is clear, you can answer our question. We guess the answer is that the IP requirement refers to the muons, and then the sentence should be revised.
- l.52: remove "four variables". Every reader is able to add 1+3.
- Prefer as is as we are introducing a list
Yes, but the Latin-numbered list is only three long. Please remove "four variables:".
- Eq.3+4: This should not be written in differential form. This makes
it inconstant (and thus wrong) with Eqs. 5-9. In this form the amplitude A should be a differential itself since the full functional form it in there (except for the decay time). However, you do not write it as such in the other equations.
- We disagree. Eq 5-9 compliment Eq. 3 &4 and are stated properly.
We still disagree with the current notation. Please consider revising.
- l.75: write "|q/p|" the same way as in the equations before.
- This would be completely incorrect! We take the absolute value to
be unity, not the phase.
We meant writing |q/p| instead of |p/q|.
- Eq.3+4: Since you make the assumptions for q/p and indirect CPV,
please simplify these equations from the start.
- It is important to show the reader what we are assuming from the
general equations
That is not the point. You can still explicitly write the assumptions and give the simplified expressions.
- l.164: You constrain the B0 mass to the PDG value for the amplitude
analysis. But in l.172 you use the signal fraction from the invariant B mass fit for the angular fit. This looks wrong. You cannot use that number after constraining to the B mass. Maybe you only constrain to the B mass for another fit, but this should be made clear in the text.
- Our procedure is consistent. The invariant mass fit gives the
signal and background fractions. We then can do whatever we wish to the events, constrain them etc... This does not change the signal and background fractions.
No, this seems to be really wrong then. As soon as you do a constrained fit on the B mass, you typically apply a chi2 cut, which changes the fractions. Even if you don't, you need to check that not too many fits failed. And if you do all this correctly, please mention this in the text to avoid any confusion.
- l.296: "efficiency-corrected" add dash.
- OK, also "background-subtracted" then
Indeed.
- Eq.34-37: These are all results, so remove the equation number.
- OK for 34 & 37, the other two are not our results.
Does not matter if it is yours or not. The EB rule is that displayed results are not numbers.
- Eq.37: C.L -> CL (without dots).
- Prefer as is
Check other papers. All of them have no dots.
Bfys-physics mailing list Bfys-physics@nikhef.nl https://mailman.nikhef.nl/mailman/listinfo/bfys-physics
Hi Vasilis,
Thanks. I disagree with his answers. I am wondering whether we should reply that we disagree with his replies. Answer like "We reconsidered but think this is OK” is not really an answer. Maybe you can say something like:
"Thanks for implemented a few of our suggestions, but we disagree with the points that you did not implement. We hope that this will be revised in the EB reading."
Otherwise the EB might think all comments are answered.
Cheers Jeroen
On 2 Apr, 2014, at 16:38 pm, V.Syropoulos <vsyropou@nikhef.nlmailto:vsyropou@nikhef.nl> wrote:
Hello all,
We got a reply from Sheldon for the second round of comments. Here is the link http://cds.cern.ch/record/1692901http://cds.cern.ch/record/1692901 I forward the original message as well.
Vasilis
-------- Original Message -------- Subject: LHCB-PAPER-2014-012-001-COMMENT-010: review received Date: Wed, 2 Apr 2014 16:20:17 +0200 From: CERN Document Server Submission Engine cds.support@cern.chmailto:cds.support@cern.ch To: vasileios.syropoulos@cern.chmailto:vasileios.syropoulos@cern.ch
Dear LHCb Colleague, Your comment LHCB-PAPER-2014-012-001-COMMENT-010 on draft LHCB-PAPER-2014-012-001 (entitled: 'Measurement of the resonant and $CP$ components in $\bar{B}^0 \to J/\psi\pi^+\pi^-$ decay') has been responded to by Sheldon Stone of the reviewers.
The review will soon be available at the following URL:
http://cds.cern.ch/record/1692901http://cds.cern.ch/record/1692901 Kind regards, CERN Document Server.
On 04/02/2014 01:26 PM, V.Syropoulos wrote: Hey Jeroen,
I just submitted your comments. https://cds.cern.ch/record/1692901
Vasilis
On 04/02/2014 10:12 AM, Jeroen Van Tilburg wrote: Hi Vasilis,
I have some follow-ups. See below. In particular, I am worried what they answer on l.164. They first extract the signal and background fractions from the B invariant mass, and then they constrain the B mass, but still use the same background and signal fractions. Does anyone know if there is a default chi2 cut in the mass-constrained fits?
Cheers Jeroen
We think you are referring to the fact that the fit best seen in Fig. 16 is somewhat above the curve at 980 MeV and slightly above around 924 MeV. We interpret these as statistical fluctuations and not indicative of any “oscillatory behavior”, since the width of the 980 is narrow enough that only about 25% of the peak number of events are present at 924 MeV (see LHCb-PAPER-2013-069 Interference of the f0(980) with other scalar states is allowed in the fit and taken into account. Adding NR improves the fit in that region, which gives quite large change on the ratio of f0(980)/f0(500). The limit is based on the NR fit, rather than the more restrictive fit without it. In addition, varying of f0(500) and f0(980) resonance parameters also changes the fit, and those are included in the syst. uncertainty. We also don’t agree that any further discussion here is warranted.
What is wrong with spending one or two sentences on this? We do not question the results. It would make the paper stronger when you say that this was carefully checked by varying the f0 resonant parameters, and that this pattern is compatible with a statistical fluctuation.
- l.5: “Cabibbo-suppressed”. Add the dash. - Not normal EB style
No. Please have a look in the EB guidelines.
- l.44: “by use of either IP requirements or detachment of the Jpsi from the primary vertex”. IP requirements of what? Of the Jpsi or its daughters? Please check. Also add meson after Jpsi. - We think the statement is clear as is and has been used previously. Adedd “meson”
If you think the statement is clear, you can answer our question. We guess the answer is that the IP requirement refers to the muons, and then the sentence should be revised.
- l.52: remove “four variables”. Every reader is able to add 1+3. - Prefer as is as we are introducing a list
Yes, but the Latin-numbered list is only three long. Please remove "four variables:".
- Eq.3+4: This should not be written in differential form. This makes it inconstant (and thus wrong) with Eqs. 5-9. In this form the amplitude A should be a differential itself since the full functional form it in there (except for the decay time). However, you do not write it as such in the other equations. - We disagree. Eq 5-9 compliment Eq. 3 &4 and are stated properly.
We still disagree with the current notation. Please consider revising.
- l.75: write “|q/p|” the same way as in the equations before. - This would be completely incorrect! We take the absolute value to be unity, not the phase.
We meant writing |q/p| instead of |p/q|.
- Eq.3+4: Since you make the assumptions for q/p and indirect CPV, please simplify these equations from the start. - It is important to show the reader what we are assuming from the general equations
That is not the point. You can still explicitly write the assumptions and give the simplified expressions.
- l.164: You constrain the B0 mass to the PDG value for the amplitude analysis. But in l.172 you use the signal fraction from the invariant B mass fit for the angular fit. This looks wrong. You cannot use that number after constraining to the B mass. Maybe you only constrain to the B mass for another fit, but this should be made clear in the text. - Our procedure is consistent. The invariant mass fit gives the signal and background fractions. We then can do whatever we wish to the events, constrain them etc… This does not change the signal and background fractions.
No, this seems to be really wrong then. As soon as you do a constrained fit on the B mass, you typically apply a chi2 cut, which changes the fractions. Even if you don’t, you need to check that not too many fits failed. And if you do all this correctly, please mention this in the text to avoid any confusion.
- l.296: “efficiency-corrected” add dash. - OK, also “background-subtracted” then
Indeed.
- Eq.34-37: These are all results, so remove the equation number. - OK for 34 & 37, the other two are not our results.
Does not matter if it is yours or not. The EB rule is that displayed results are not numbers.
- Eq.37: C.L -> CL (without dots). - Prefer as is
Check other papers. All of them have no dots.
_______________________________________________ Bfys-physics mailing list Bfys-physics@nikhef.nlmailto:Bfys-physics@nikhef.nl https://mailman.nikhef.nl/mailman/listinfo/bfys-physics
_______________________________________________ Bfys-physics mailing list Bfys-physics@nikhef.nlmailto:Bfys-physics@nikhef.nl https://mailman.nikhef.nl/mailman/listinfo/bfys-physics
-
Jeroen Van Tilburg -
V.Syropoulos