[evla-sw-discuss] Amplitudes greater than 1 in TelCal

[Bryan Butler]

well, a correlated power is still a correlation, but i get the point.  i 
just misunderstood the normalization (or lack thereof).

	-bryan


Ken Sowinski wrote, On 11/1/11 15:42 PM:
> On Tue, 1 Nov 2011, Bryan Butler wrote:
>
>>
>> it's still a correlation, and it's not scaled (or at least i *think*
>> it's not scaled), so should always be <= 1.0.
>
> As MIchael said, it is not a correlation but a "correlated power".
> Autocorrelation come out to be about 2.7**2 if everything is set right.
>
> Example: 3c273 has about 0.1 correlation at many bands and the fringe
> disaply provedes numbers between 0.5 and 0.8 for most antennas.
>
> The number would more resemble a correlation coefficient if devided
> the the requeanter RMS sqaured.
>
> Ken
>
>
>> Michael Rupen wrote, On 11/1/11 15:29 PM:
>>> But this is NOT a raw correlation coefficient -- why should it be <=1 ?
>>>
>>> Michael
>>>
>>>>
>>>> it should never be > 1.0, since it is a correlation.
>>>>
>>>> -bryan
>>>>
>>>>
>>>> Keith Cummings wrote, On 11/1/11 15:02 PM:
>>>>> Michael caused problems today by running Lsysstart.55866.792821435185
>>>>> which produced amplitudes greater than 1. TelCal's Antsol routine does
>>>>> not approve of amplitudes (actually amplitudes squared) greater than 1
>>>>> or less than 1e-20. If five integrations within a
>>>>> baseline-baseband-polarization-subband are out-of-range, antsol marks
>>>>> this data as flagged for both antennas in this baseline
>>>>> (antenna-baseband-polarization).
>>>>>
>>>>> The low side (less than 1e-20) is probably fine. Michael is suggesting
>>>>> that the high side restriction (greater than 1) might need to be
>>>>> removed.
>>>>>
>>>>> Thoughts?
>>>>>
>>>>> -Keith
>>>>> _______________________________________________
>>>>> evla-sw-discuss mailing list
>>>>> evla-sw-discuss at listmgr.cv.nrao.edu
>>>>> http://listmgr.cv.nrao.edu/mailman/listinfo/evla-sw-discuss
>>>> _______________________________________________
>>>> evla-sw-discuss mailing list
>>>> evla-sw-discuss at listmgr.cv.nrao.edu
>>>> http://listmgr.cv.nrao.edu/mailman/listinfo/evla-sw-discuss
>>>>
>> _______________________________________________
>> evla-sw-discuss mailing list
>> evla-sw-discuss at listmgr.cv.nrao.edu
>> http://listmgr.cv.nrao.edu/mailman/listinfo/evla-sw-discuss
>>