[R] Transfer function observed vs predicted values graph problem
Hi, I am trying to make a palaeoenvironmental transfer function using the R package rioja that predicts the water-table (measured as depth to the water table) of an area given the testate amoebae that are found there. I've carried out weighted averaging of the data and am trying to produce a graph that shows the observed water-table versues the model's predicted values. Following the instructions in the rioja help booklet (see below), I end up with a graph where the origin is not at the bottom left of the diagram, i.e. the graph is showing some values that suggest that the water table is, say, 1m above ground. I've tried entering the water-tables as negative values but the same thing happens. Does anybody know if there something I'm missing out? Or is there a way that, if the values returned are less than 0, then they can automatically be put just as 0? Any help would be most appreciated, Thank you, Matthew My environmental matrix (x) is: SampleId WTD Moisture pH EC 1 1 20 91.72700 3.496674 85.02688 2 2 2 93.88913 3.550794 85.69465 3 3 26 90.30269 3.948559 113.19206 4 4 5 94.14427 3.697213 48.56375 5 5 30 90.04269 3.745020 108.57278 90 GAL_15 70 94.07849 3.777932 66.77673 The species matrix (y) contains the abundance of 32 species over 90 sites, set out like this F1 AmpFlav AmpWri ArcCat ArcDis 1 1 22.2929936 0.000 0.000 0.000 2 2 30.9677419 0.000 0.000 3.2258065 fit - WA(y, x, tolDW = FALSE, use.N2=TRUE, check.data=TRUE, lean=FALSE) # plot predicted vs. observed plot(fit) plot(fit, resid=TRUE) # Water-table reconstruction pred - predict(fit, y) #plot the reconstruction plot(sites, pred$fit[, 1], type=b) # cross-validation model using bootstrapping fit.xv - crossval(fit, cv.method=boot, nboot=1000) par(mfrow=c(1,2)) plot(fit) plot(fit, resid=TRUE) plot(fit.xv, xval=TRUE) plot(fit.xv, xval=TRUE, resid=TRUE) -- View this message in context: http://r.789695.n4.nabble.com/Transfer-function-observed-vs-predicted-values-graph-problem-tp3320663p3320663.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Rioja package, creating transfer function, WA, Error in FUN
Dear Peter, Thank you very much for your suggestion, I went through the matrices and removed the headings for pH, WTD etc and did the same for the site names, which I made purely numerical, and now the WA function is working, Thanks again! Matthew On Fri, Feb 11, 2011 at 12:44 AM, Peter Ehlers [via R] ml-node+3300569-1208751190-212...@n4.nabble.com wrote: On 2011-02-10 09:40, mdc wrote: Hi, I am a new R user and am trying to construct a palaeoenvironmental transfer function (weighted averaging method) using the package rioja. I've managed to insert the two matrices (the species abundance and the environmental data) and have assigned them to the y and x values respectively. When I try and enter the 'WA' function though, I get an 'Error in FUN' message (see below for full values). Alas, I do not know what this means and have struggled to find similar problems to this online. Is there a step I've missed out between assigning the matrices and the WA function? SWED=odbcConnectExcel(file.choose()) (SWED is the environmental data file) sqlTables(SWED) Env=sqlFetch(SWED, Sheet1) odbcClose(SWED) Env SampleId WTD Moisture pH EC 1 N1_1 20 91.72700 3.496674 85.02688 2 N1_22 93.88913 3.550794 85.69465 3 N1_3 26 90.30269 3.948559 113.19206 4 N1_45 94.14427 3.697213 48.56375 5 N1_5 30 90.04269 3.745020 108.57278 90 GAL_15 70 94.07849 3.777932 66.77673 STEST=odbcConnectExcel(file.choose()) sqlTables(STEST) (STEST is the species abundance file) Spe=sqlFetch(STEST, Sheet8) odbcClose(STEST) Spe (The species data contains the abundance of 32 species over 90 sites, set out like this) F1AmpFlavAmpWri ArcCat ArcDis 1N1_1 22.2929936 0.000 0.000 0.000 2N1_2 30.9677419 0.000 0.000 3.2258065 library(rioja) y-as.matrix(Spe) x-as.matrix(Env) WA(y, x, tolDW = FALSE, use.N2=TRUE, check.data=TRUE, lean=FALSE) (the command from the WA section of the rioja booklet) Error in FUN(newX[, i], ...) : invalid 'type' (character) of argument Well, the error message is fairly clear: you're feeding in something of type 'character' where something else (presumably) numeric is wanted. I don't use rioja, but a quick glance at the documentation for WA shows that x should be 'a vector of environmental values to be modelled'. The example uses pH which is almost surely not a character vector. Your x is a *matrix* of *character* values. Possibly, you want to pull, say, pH out of your Env, convert to numeric and try that. Ditto for the other variables. If the above is total nonsense, please forgive my rioja ignorance and wait for more cogent advice from someone more knowledgeable than I. Peter Ehlers Any help would be most appreciated, Best wishes, Matthew __ [hidden email] http://user/SendEmail.jtp?type=nodenode=3300569i=0mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- If you reply to this email, your message will be added to the discussion below: http://r.789695.n4.nabble.com/Rioja-package-creating-transfer-function-WA-Error-in-FUN-tp3299636p3300569.html To unsubscribe from Rioja package, creating transfer function, WA, Error in FUN, click herehttp://r.789695.n4.nabble.com/template/NamlServlet.jtp?macro=unsubscribe_by_codenode=3299636code=bWF0dC5kLmNvZUBnbWFpbC5jb218MzI5OTYzNnwyMDMzMDk2ODg2. -- View this message in context: http://r.789695.n4.nabble.com/Rioja-package-creating-transfer-function-WA-Error-in-FUN-tp3299636p3302063.html Sent from the R help mailing list archive at Nabble.com. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Rioja package, creating transfer function, WA, Error in FUN
Hi, I am a new R user and am trying to construct a palaeoenvironmental transfer function (weighted averaging method) using the package rioja. I've managed to insert the two matrices (the species abundance and the environmental data) and have assigned them to the y and x values respectively. When I try and enter the 'WA' function though, I get an 'Error in FUN' message (see below for full values). Alas, I do not know what this means and have struggled to find similar problems to this online. Is there a step I've missed out between assigning the matrices and the WA function? SWED=odbcConnectExcel(file.choose()) (SWED is the environmental data file) sqlTables(SWED) Env=sqlFetch(SWED, Sheet1) odbcClose(SWED) Env SampleId WTD Moisture pH EC 1 N1_1 20 91.72700 3.496674 85.02688 2 N1_22 93.88913 3.550794 85.69465 3 N1_3 26 90.30269 3.948559 113.19206 4 N1_45 94.14427 3.697213 48.56375 5 N1_5 30 90.04269 3.745020 108.57278 90 GAL_15 70 94.07849 3.777932 66.77673 STEST=odbcConnectExcel(file.choose()) sqlTables(STEST) (STEST is the species abundance file) Spe=sqlFetch(STEST, Sheet8) odbcClose(STEST) Spe (The species data contains the abundance of 32 species over 90 sites, set out like this) F1AmpFlavAmpWri ArcCat ArcDis 1N1_1 22.2929936 0.000 0.000 0.000 2N1_2 30.9677419 0.000 0.000 3.2258065 library(rioja) y -as.matrix(Spe) x -as.matrix(Env) WA(y, x, tolDW = FALSE, use.N2=TRUE, check.data=TRUE, lean=FALSE)(the command from the WA section of the rioja booklet) Error in FUN(newX[, i], ...) : invalid 'type' (character) of argument Any help would be most appreciated, Best wishes, Matthew -- View this message in context: http://r.789695.n4.nabble.com/Rioja-package-creating-transfer-function-WA-Error-in-FUN-tp3299636p3299636.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.