svr.jl link fix

examples/svr/svr.jl

@@ -1,44 +1 @@
-import Mads
-
-@info("Support Vector Regression (SVR) analysis")
-
-Random.seed!(2017)
-
-md = Mads.loadmadsfile(joinpath("..", "models", "internal-polynomial-model", "internal-polynomial.mads"))
-rootname = Mads.getmadsrootname(md)
-svrexec, svrread, svrsave, svrclean = Mads.makesvrmodel(md, 100)
-
-numberofsamples = 100
-paramdict = Mads.getparamrandom(md, numberofsamples)
-paramarray = hcat(map(i->collect(paramdict[i]), keys(paramdict))...)
-predictions = Mads.forward(md, paramdict)'
-
-Mads.madsinfo("Model predictions ...")
-Mads.spaghettiplot(md, predictions, keyword="w13a-model", format="PNG")
-Mads.display("$rootname-w13a-model-$numberofsamples-spaghetti.png")
-
-@time svrpredictions = svrexec(paramarray)
-@info("SVR discrepancy $(maximum(abs.(svrpredictions .- predictions)))")
-
-Mads.madsinfo("SVR predictions ...")
-Mads.spaghettiplot(md, svrpredictions, keyword="w13a-svr", format="PNG")
-Mads.display("$rootname-w13a-svr-$numberofsamples-spaghetti.png")
-
-sa = Mads.efast(md)
-Mads.plotobsSAresults(md, sa, format="PNG")
-Mads.display("$rootname-efast-385.png")
-
-mdsvr = deepcopy(md)
-mdsvr["Julia model"] = svrexec
-mdsvr["Filename"] = "$rootname-svr.mads"
-
-sasvr = Mads.efast(mdsvr)
-Mads.plotobsSAresults(mdsvr, sasvr, format="PNG")
-Mads.display("$rootname-svr-efast-385.png")
-
-svrsave()
-svrclean()
-svrread()
-svrclean()
-
-Mads.rmdir("svrmodels")
+runtest.jl