The ATCA Forum

Hi,

I'm attempting to get a flux reading of my visibility data at 33GHz and 35GHz using uvfit but I keep receiving a buffer overflow error. I've read the previous forum post on this (https://atcaforum.atnf.csiro.au/viewtop ... rflow#p385). However, I'm running on my university's supercomputer OZSTAR and I believe memory shouldn't be an issue. Is there a way to increase the buffer limit without recompiling?

Cheers,
Brodie

Hi Brodie,

No, you must recompile Miriad to change the buffer limit.

Have you tried uvsfit instead of uvfit?

Hi Brodie,

unless you are fitting sources far from the phase centre and bandwidth smearing is an issue,
it is quicker to just use the line parameter to do some spectral averaging of the data: You are allowed 20 million visibilities, so just average things down to get below that limit.
You can try a few settings to convince yourself it makes little difference to the result.

uvsfit has a slightly higher limit, 35 million visibilities.

Cheers,

Mark

Thank you both for the advice