maxamp, maxphs Set limits on acceptable amplitude and phase corrections in selfcal. EXAMPLES -------- 1. To prevent selfcal from making amplitude corrections of more than 1.1 (unitless scale factor), or phase corrections of more that 10 degrees, type: 0>selflims 1.1,10 Selfcal amplitude corrections will be limited to 0.909091 -> 1.1. Selfcal phase corrections will be limited to -10 -> 10 degrees. 0> All but the first line are the response from selflims confirming the new limits. 2. To display the currently selected limits omit all arguments: 0>selflims Selfcal amplitude corrections will be limited to 0.909091 -> 1.1. Selfcal phase corrections will be limited to -10 -> 10 degrees. 0> 3. To cancel all limits type: 0>selflims 0,0 Selfcal amplitude corrections will not be limited. Selfcal phase corrections will not be limited. 0> PARAMETERS ---------- maxamp - If maxamp < 1 then no limit will be placed on amplitude corrections. Otherwise selfcal solution intervals that result in one or more amplitude corrections outside the range 1/maxamp to maxamp will be left un-corrected. maxphs - If maxphs <= 0 then no limit will be placed on amplitude corrections. Otherwise selfcal solution intervals that result in one or more phase corrections outside the range -maxphs to maxphs degrees will be left un-corrected. When a new data-set is read in using the 'observe' command the limits are reset to provide no limits on corrections. CONTEXT ------- In some data sets where the data is sparse, noisy or not well constrained for other reasons, selfcal may produce ridiculously large aberrant corrections. The selflims command is used before calling selfcal to set limits on the allowed corrections. Note that rejection of the solution for a given solution bin does NOT result in the effected visibilities being flagged - regardless of the flagging disposition set by any previous use of the selfflag command. RELATED COMMANDS ---------------- selfcal - Perform self-calibration. selfflag - Used to control the fate of un-correctable data. selftaper - Used to down-weight short baselines during selfcal. selfant - Set antenna based constraints in selfcal.