MHO_MinWeight

Purpose

MHO_MinWeight implements a weight-based selection filter that zeroes out all weight values below a user-specified minimum threshold. This operator effectively removes low-quality visibility data points from downstream processing by driving their weights to zero.

Control File Trigger

• Keyword: min_weight

• Category: selection

• Priority: 3.5

Parameters for min_weight

Parameter	Type	Description
value	real	Minimum weight threshold; all weights below this value are set to zero.

Input Data

This operator acts on the weight_type container.

Algorithm

MHO_MinWeight has no Initialize method; all work occurs in ExecuteInPlace.

The builder (MHO_MinWeightBuilder) retrieves the weight container from the container store, and skips construction entirely if the configured value is zero.

Execution (``ExecuteInPlace``):

1. Iterate over every element in the weight array.

2. For each weight w, compare against the threshold: if w < w_min, w = 0.

Effect on Data

The operator modifies the weight_type container in place. All weight values strictly less than the configured minimum threshold are set to zero. Zeroed weights cause downstream operators to exclude the corresponding visibility data points from fringe-fitting, averaging, or any weighted computation.