There can be some pathalogical cases where the SNR increases in spite of large increases in the RMS. This is ok if we have reason to believe that the flagging, which leads to the increase in the RMS, is warranted (e.g. we expect high SNR solutions but aren't getting them). But in the event that it may not be warranted (e.g. we are in the low SNR regime to begin with), it is better to fail the solint than to accept the significant RMS increase.
This PR adds a check for RMS increases, coupled with a low expected SNR, as a reason to fail a solint.
There can be some pathalogical cases where the SNR increases in spite of large increases in the RMS. This is ok if we have reason to believe that the flagging, which leads to the increase in the RMS, is warranted (e.g. we expect high SNR solutions but aren't getting them). But in the event that it may not be warranted (e.g. we are in the low SNR regime to begin with), it is better to fail the solint than to accept the significant RMS increase.
This PR adds a check for RMS increases, coupled with a low expected SNR, as a reason to fail a solint.