nanmax

FORMAT: [Y,IDX] = NANMAX(A,DIM,[B])

  Maximum ignoring NaNs

  This function enhances the functionality of NANMAX as distributed in
  the MATLAB Statistics Toolbox and is meant as a replacement (hence the
  identical name).

  If fact NANMAX simply rearranges the input arguments to MAX because
  MAX already ignores NaNs.

  NANMAX(A,DIM) calculates the maximum of A along the dimension DIM of
  the N-D array X. If DIM is omitted NANMAX calculates the maximum along
  the first non-singleton dimension of X.

  NANMAX(A,[],B) returns the minimum of the N-D arrays A and B.  A and
  B must be of the same size.

  Comparing two matrices in a particular dimension is not supported,
  e.g. NANMAX(A,2,B) is invalid.

  [Y,IDX] = NANMAX(X,DIM) returns the index to the maximum in IDX.

  Similar replacements exist for NANMIN, NANMEAN, NANSTD, NANMEDIAN and
  NANSUM which are all part of the NaN-suite.

Syntax

[y,idx] = nanmax(a,dim,b)

See also