fixed findbugs warnings and added more explanation about the corresponding behavior in javadoc

git-svn-id: https://svn.apache.org/repos/asf/commons/proper/math/trunk@762136 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Luc Maisonobe 2009-04-05 19:10:44 +00:00
parent 5acb8b11b5
commit 28738e0405
3 changed files with 15 additions and 0 deletions

View File

@ -84,6 +84,11 @@
</Or>
<Bug pattern="EI_EXPOSE_REP" />
</Match>
<Match>
<Class name="org.apache.commons.math.ode.sampling.MultistepStepInterpolator"/>
<Method name="reinitialize" params="double[],double[],double[][],boolean" returns="void" />
<Bug pattern="EI_EXPOSE_REP2" />
</Match>
<Match>
<Class name="org.apache.commons.math.linear.BigMatrixImpl"/>
<Method name="&lt;init>" params="java.math.BigDecimal[][],boolean" returns="void" />

View File

@ -93,6 +93,8 @@ public abstract class MultistepStepInterpolator
}
/** Reinitialize the instance
* <p>Beware that all arrays <em>must</em> be references to integrator
* arrays, in order to ensure proper update without copy.</p>
* @param y reference to the integrator array holding the state at
* the end of the step
* @param previousT reference to the integrator array holding the times

View File

@ -32,6 +32,14 @@ import org.apache.commons.math.ode.SecondOrderIntegrator;
* handlers can use these objects to retrieve the state vector at
* intermediate times between the previous and the current grid points
* (this feature is often called dense output).</p>
* <p>One important thing to note is that the step handlers may be so
* tightly bound to the integrators that they often share some internal
* state arrays. This imply that one should <em>never</em> use a direct
* reference to a step interpolator outside of the step handler, either
* for future use or for use in another thread. If such a need arise, the
* step interpolator <em>must</em> be copied using the dedicated
* {@link #copy()} method.
* </p>
*
* @see FirstOrderIntegrator
* @see SecondOrderIntegrator