Re: svn commit: r739504 - in /commons/proper/math/trunk/src: java/org/apache/commons/math/linear/SparseRealVector.java test/org/apache/commons/math/linear/SparseRealVectorTest.java

2009-01-31 Thread Bill Barker
It's a straight copy/paste from RealVectorImpl. I agree that public isn't best (but would probably go for protected instead of private so still usable by subclasses). However, making it non-public breaks the junit tests (which are a copy of the RealVectorImpl junit test with a search/replace,

Re: svn commit: r739504 - in /commons/proper/math/trunk/src: java/org/apache/commons/math/linear/SparseRealVector.java test/org/apache/commons/math/linear/SparseRealVectorTest.java

2009-01-31 Thread Rahul Akolkar
Minor comment, and arguably subjective -- I find the signature for checkVectorDimensions(int) a little odd for a public method (I'd make it private). -Rahul On Sat, Jan 31, 2009 at 10:13 AM, wrote: > Here are my comments. > > The two isZero methods seem inconsistent with each other. If I expli

[lang] EqualsBuilder

2009-01-31 Thread Peter Henderson
Hi Folks, hopefully this question is appropriate for the dev list. I have been benchmarking EqualsBuilder and run into what must be a common gotcha. The docs for EqualsBuilder suggests that one construct the equals method:- @Override public boolean equals(Object obj) { if (ob