jun12-dec12: sysperf, pnt error

Calibration data from jun12 through dec12 was used to look at the system performance and pointing errors of various receivers.

History:

19oct12 - jennings relay from 430 dome polB removed. Prior to this, little power was getting into polB.

For each receiver, calibration data for the specified range was extracted from the datasets.
calibration scans were excluded if:

they had an exceptionally large pointing error for this receiver (say > 60 asecs)
the computed gain was very low (< 1 K/Jy)
These were to throw out scans with problems.

For each receiver 3 sets of plots were made:

Table of plots:

Rcv calibration and pointing error plots	avg pointing errors Asecs				Updated model offsets 13jan13
	za errors		azimuth errors
	mean	rms	mean	rms
327 (.ps) (.pdf)	9.9	12.5	4.1	10.6	yes
430Dome (.ps) (.pdf)	22.3	8.6	25.4	8.7	yes
lbw (.ps) (.pdf)	4.1	4.0	.07	6.7	yes
sbn (.ps) (.pdf)	-11.3	3.9	-9.7	4.9	yes
sbw (.ps) (.pdf)	9.0	3.1	6.2	5.6	yes
cb (.ps) (.pdf)	3.8	4.0	2.8	5.2	yes
cbh (.ps) (.pdf)	2.5	4.5	1.3	4.5	no
xb (.ps) (.pdf)	.8	4.1	-.1	7.1	no

The offset is normally computed from the model. This uses sbn and 20 sources covering the entire dish. The offsets for the other receivers are then computed relative to sbn using 2 or 4 sources.This was done back in apr-may12.
With the installation of the new sbn receiver, the standard model offset (from the 20 sources) is no longer valid for the sbn receiver. So it's offsets had to be updated from the sources after the installation.
I decided to update the other receiver offsets so everything changes at the same epoch (even though there is not complete dish coverage).

processing: ~x101/130112/pnter.pro

 home_~phil