<div dir="ltr"><div>Dear All,</div><div><br></div><div>Just a quick question on whether there should be a performance difference between different approaches in moving data from a local Vec to a global Vec:</div><div><br></div>
<div><div>qbc is a multi-dimensional, Fortran ordered numpy array</div></div><div><br></div><div><div># METHOD 1</div><div><br></div><div><div># with fset(self,q) defined as follows:</div><div><div> def fset(self,q):</div>
<div> self.gqVec.setArray(q.reshape([-1], order = 'F'))</div></div></div><div><br></div><div> self.q = qbc[:,mbc:-mbc,mbc:-mbc]</div><div><br></div></div><div># METHOD 2 </div><div><br></div><div>
self.lqVec.placeArray(qbc.reshape([-1], order = 'F'))</div><div> self.q_da.localToGlobal(self.lqVec,self.gqVec)</div><div> self.lqVec.resetArray() </div><div><br></div><div><div># METHOD 3 </div>
<div><br></div><div> self.lqVec.placeArray(qbc)</div><div> self.q_da.localToGlobal(self.lqVec,self.gqVec)</div><div> self.lqVec.resetArray() </div></div><div><br></div><div>A</div></div>