Last update: 12 feb98
RUN 014
19 December 97. A run using MC data sets from Bernd. Here are the inputs:
singles pi+,
pi-,
K+,
K-,
and pairs pi+,
pi-,
K+,
K-, and these were the
input parameters. This run was with no
systematic errors, and resulted in a bad chi**2. See the log file.
RUN 015
19 December 97. Same as run 14, but with added 10% systematic errors,
see the input parameter file.
See the log file.
(note that the input files had an implicit efficiency of 90% in it.
RUN 016
6 Jan 1998. Same as run 15, but to compensate for the implicit efficiency, the
input Z and A are reduced by a factor of 0.9.
See the input parameter file.
See the log file. This run took a long time;
normally, 15-20 iterations are done, but this one took 57 rounds. The
chisq/dof overall is 24, not too good. If you look at the log file, you can
see that this is completely dominated by the terrible fit to the pion BE data,
also reflected in the huge R found.
Here is the BEST point !
**********************************************
Total chi^2 = 59739.75630549779
chi^2/d.o.f = 24.61465031128875
Probability of perfect model = .0
Fixed parameters:
aT = -.9900000
lamK = 1.0000000
normpi= 1.0000000
normK = 1.0000000
Variable parameters:
T = 98.8762643 +/- 3.2403190 MeV
vT/c = .8254691 +/- .0197095
et0 = 2.6069185 +/- .0251666
muB/T = 3.9538487 +/- .2574856
tau = .0153866 +/- .0017893 fm/c
R = 98.8611252 +/- 1.7384788 fm
ycm = -.0100097 +/- .0110566
fs = 1.3466445 +/- .0991520
lam = .2105840 +/- .0073149
Calculated Quantities:
vs/c = -.0100094 + .0110563 -.0110539
vL/c = .9891774 + .0005285 -.0005555
muB = 390.9417876 + 15.3943255 -16.3575785 MeV
t1 = .0015387 + .0001779 -.0001782 fm/c
t3 = .1048669 + .0140059 -.0136073 fm/c
t2-t1 = .0138479 + .0016007 -.0016036 fm/c
dtau = .0145181 + .0017016 -.0016947 fm/c
z3 = .1037320 + .0129497 -.0134400 fm
muS = 32.0679171 + 3.2786466 -3.0755850 MeV
muI = -2.9707652 + .3574922 -.3846667 MeV
Bar# = 10.4388677 + 1.4051798 -1.2460073
Bproj = -.0113163 + .0125133 -.0126092
Btarg = 10.4501840 + 1.4091967 -1.2484270
brdn1 = .00727418 + .00164945 -.00143050 fm^-3
brdn2 = .00726493 + .00164859 -.00142927 fm^-3
---------------------------------------------------------
chi^2/dof for each data set ...
335 333.762315270936 365.2731403196502 1.09441097333932
335 333.762315270936 527.0940649252853 1.57924978587654
257 256.0504926108375 400.3445204002943 1.56353739576188
213 212.2130541871921 870.4608626349222 4.10182524335705
324 322.8029556650246 28304.91590757799 87.68481022506574
324 322.8029556650246 28306.71355607188 87.69037909753838
324 322.8029556650246 514.880639684714 1.59503074754684
324 322.8029556650246 450.0736138830471 1.3942673261954
chi^2 = 59739.75630549779 chi^2/dof = 24.61465031128875
chi^2_min = 59739.756305497789
RUN 017
7 Jan 1998. Same as run 16, but the input singles data are now restricted to
rapidity -1>y>+1. This is done because this particular EOS leads to dn/dy
distributions which have a central peak and pronounced shoulders. FREEZER has
no shoulders, and therefore should have trouble with the large-y data.
The input Z and A are reduced by a factor of 0.9, as before.
Input files are:
singles pi+,
pi-,
K+,
K-,
and pairs pi+,
pi-,
K+,
K-
See the input parameter file.
See the log file.
RUN 018
10 Jan 1998. Same as run 17, but with R=8.0 fm and fixed. This was because we
saw in run 17 that the radius had run away.
See the input parameter file.
See the log file.
The chisq/ndf went up, as expected, to 66, but T is now very large,
compensated by a very low v. Apparently, what was found was just another point
in the T-v valley of low chisq.
RUN 019
12 Jan 1998. Same as run 18, but only fitting to the singles.
See the input parameter file.
See the log file.
RUN 020
13 Jan 1998. Same as run 19, but R is free and systematic errors set to 0
(note the individual point already have a 5% error on them).
See the input parameter file.
See the log file.
RUN 021
14 Jan 1998. Same as run 20, but T is fixed at 140. Curent settings:
- T fixed at 140 meV
- Singles spectra only
- -1 < y < +1
- systematic errors set to 0.0
- Nuclei are 90% of nominal, to compensate for 90% 'experimental efficiency'.
(note the individual point already have a 5% error on them).
See the input parameter file.
See the log file.
RUN 022
15 Jan 1998. Same as run 21, but add the kaon correlations
- T fixed at 140 meV
- Singles pi, K, K correlations
- -1 < y < +1
- systematic errors set to 0.0
- Nuclei are 90% of nominal, to compensate for 90% 'experimental efficiency'.
- NOTE THAT ALL THIS TIME AT WAS FIXED AT -0.99
(note the individual point already have a 5% error on them).
See the input parameter file.
See the log file.
RUN 023
26 Jan 1998. A new data set: the 'efficiency' is now 100%, and there are more
integration points in the pairs files. This set is limited to -1< y < +1.
Input files are:
singles pi+,
pi-,
K+,
K-,
and pairs pi+,
pi-,
K+,
K-
All parameters are free for this run.
See the input parameter file.
and the output log file.
aT runs away to +1, and the program stops.
RUN 024
****************** ignore everything below this line *******************
15 Jan 1998. Same as run 23, but with aT fixed at +0.99, as per instruction
from FREEZER. Terrible fit. R goes off to infinity, and tau goes to 0.
See the input parameter file.
and the output log file.
RUN 025
20 Jan 1998. Same as
See the input parameter file.
and the output log file.
RUN 026
2 feb 1998. T=140, aT=-1, tau=
See the input parameter file.
and the output log file.
RUN 027
3 feb 1998. Singles only, T fixed at 140, aT fixed at -1.
See the input parameter file.
and the output log file.
OK run found reasonable v.
RUN 028
5 feb 1998. Singles only, let T loose, aT fixed at -1.
See the input parameter file.
and the output log file.
T floated down from 140 to 116.
RUN 029
6 feb 1998. Singles only, everything free
See the input parameter file.
and the output log file.
Basically, it was clear we could not get a good fit to all the data sets. At a
meeting on 11 feb 98, Ray suggested we should increase the number of
integration points. This we did, and the results are on
the next page.