<div><br></div><div><meta http-equiv="content-type" content="text/html; charset=utf-8"><br class="Apple-interchange-newline">2011/3/22 Barry Smith <span dir="ltr"><<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>></span><br>
<blockquote class="gmail_quote" style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0.8ex; border-left-width: 1px; border-left-color: rgb(204, 204, 204); border-left-style: solid; padding-left: 1ex; ">
<div class="im"><br><br></div> VecView() for DM global vectors present the vector entries in the global natural ordering. Since PETSc divides the domain with DA into squareish blocks when you print the values on a process (by calling DAVecGetArray) it is not in the global natural ordering. Hence the two are different.<br>
<font color="#888888"><br> Barry<br></font><div><div></div></div></blockquote></div><br>Thank You for you answer! Yes now I can see from the output, that VecView uses natural ordering, but its also very importaint to understand where (on which processor) elements are finaly stored. If I got it in a right way, VecView uses natural ordering and gives the WRONG information about actual distribution of memory among processors. And elemens are really splitted according the DA implementation. Am I right? <div>
<br></div><div>Alexey</div>