Hi, I think some method names that represent operators are wrong, at least inconsistent by naming and/or by implementation. I opend an issue: https://issues.apache.org/jira/browse/MATH-643
Comming from matlab/octave the octave operator .+ is specified ad element-by-element addition. http://www.gnu.org/software/octave/doc/interpreter/Arithmetic-Ops.html#Arithmetic-Ops If one argument is a scalar this will be applied to all entries of the matrix (here vector). if the if both are matricies (here vectors) the dimensions must agree, then all entries are added element-by-element. mapAdd() is for me confusing. The method map(), which maps a function to each entry is OK, but what is mapAdd() in an interface? - In the implementation of Abstractrealvector.mapAddToSelf() is the class Add as function call used, but this is only relevant for the implementation and should not leak into the interface. So add() should become ebeAdd(). Or add ebeAdd()? For all add() operands addToSelf() is missing - mapAdd() provides this. Here is the current naming: RealVector add(double[] v); RealVector add(RealVector v); RealVector mapAdd(double d); RealVector mapAddToSelf(double d); RealVector ebeMultiply(RealVector v); RealVector ebeMultiply(double[] v); RealVector mapMultiply(double d); RealVector mapMultiplyToSelf(double d); This should be renamed to: RealVector ebeAdd(double[] v); RealVector ebeAddToSelf(double[] v); RealVector ebeAdd(RealVector v); RealVector ebeAddToSelf(RealVector v); RealVector ebeAdd(double d); RealVector ebeAddToSelf(double d); RealVector ebeMultiply(RealVector v); RealVector ebeMultiplyToSelf(RealVector v); RealVector ebeMultiply(double[] v); RealVector ebeMultiplyToSelf(double[] v); RealVector ebeMultiply(double d); RealVector ebeMultiplyToSelf(double d); Arne --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
