<div dir="ltr"><pre class="" id="yui_3_16_0_1_1415521224923_97661" style="font-family:'Courier New';white-space:pre-wrap;color:rgb(0,0,0);font-size:12.7272720336914px"><span class="" id="yui_3_16_0_1_1415521224923_97660" style="font-style:italic">Dear all,<br>I am using the package SpatioTemporal and trying to analyze PM data.When I use estimate function I receive this message:Lapack routine dgesv: system is exactly singular: U[1,1] = 0<br class=""><br class="">> est.EBOD.model <- estimate(EBOD.model, x.init,<br class="">+ type="p", hessian.all=FALSE)<br class="">Optimisation using starting value 1/2<br class="">N = 9, M = 5 machine precision = 2.22045e-16<br class="">At X0, 0 variables are exactly at the bounds<br class="">At iterate 0 f= 31227 |proj g|= 15<br class="">At iterate 10 f = 3036.8 |proj g|= 2.4429<br class=""><br class="">iterations 17<br class="">function evaluations 31<br class="">segments explored during Cauchy searches 21<br class="">BFGS updates skipped 0<br class="">active bounds at final generalized Cauchy point 0<br class="">norm of the final projected gradient 0.0555717<br class="">final function value 3036.35<br class=""><br class="">F = 3036.35<br class="">final value 3036.345750 <br class="">converged<br class=""><br class="">Error in diag(solve(res[[i]]$hessian)) : <br class=""> error in evaluating the argument 'x' in selecting a method for function 'diag': Error in solve.default(res[[i]]$hessian) : <br class=""> Lapack routine dgesv: system is exactly singular: U[1,1] = 0<br class=""><br class="">I think this error occur due to high collinearity between stations but I have
checked the correlations and they are not high:<br class=""> | PM31 PM32 PM33 PM34 PM35 PM36<br class="">-------------+------------------------------------------------------<br class=""> PM31 | 1.0000<br class=""> PM32 | 0.3547 1.0000<br class=""> PM33 | 0.6414 -0.1596 1.0000<br class=""> PM34 | 0.7102 0.1937 0.6225 1.0000<br class=""> PM35 | 0.2578 0.6152 -0.2569 0.3733 1.0000<br class=""> PM36 | 0.4386 0.6826 0.2227 0.2510 0.1930 1.0000<br class=""><br class="">I have attached R code with necessary files. <br class="">If it is possible please help me to understand the problem and find a solution for this problem. <br class="">Thanks for your help in advance. <br class="">Kindest regards, <br class="">Mahboubeh</span></pre></div>