Hi Heidi, That's *exactly* what i wanted- an example of how to adjust these major aspects of a heat map, which i hope will help biologists in the future attempting this. Additionally the actual metrics are precisely what my data appear to 'like'. But i still cant move over to heatmap.2 having discovered the usefulness of heatplot's "dualScale=" argument which really 'tames' one or two outlying samples in my data (but may ultimately be discarded). And the "zlim=" arg is also very convenient. Both absent in heatmap.2 and too complex for me to 'port' them to heatmap.2 . I did also successfully mod heatplot() to use any 3 colors by replacing the following line in heatplot: hmcol <- colorRampPalette(brewer.pal(10, "RdBu"))(256) with, eg: hmcol <- redgreen(256) #To yield the red/black/green or, per your suggestion: hmcol <- colorpanel(256, "blue", "black", "yellow") So for now, its heatplot with the unwanted row dendrogram, or heatmap without the additional row & column scaling. Thanks again for your help, really appreciated, karl On 8/8/2010 9:51 PM, Heidi Dvinge wrote: >> Hi Heidi, >> >> Thank you for your thoughts and suggestions, >> >> On 8/7/2010 6:49 PM, Heidi Dvinge wrote: >>> Hello Karl, >>> >>> I'm not familiar with the made4 package, but it looks like some of the >>> changes you want cannot override the defaults from heatplot. Are there >>> any >>> particular reason why you don't want to use heatmap.2 directly? Using >>> your >>> example from below, then >> >> The stated defaults of heatplot() follow the methodology by Eisen et al >> 1998. This is my starting point, and i must admit, my data 'looks' v. >> good using these distance (Pearson Correlation coefficient) and linkage >> (average) methods. >> > Is something like this more what you're after? > > library(gplots) > my.mat<- matrix(rnorm(250), nrow=50, ncol=5) > dist.pear<- function(x) as.dist(1-cor(t(x))) > hclust.ave<- function(x) hclust(x, method="average") > > heatmap.2(my.mat, col=c(redgreen(256)), dendrogram="col", trace="none", > density.info="none", distfun=dist.pear, hclustfun=hclust.ave) > > Although, looking at the heatplot function they use a slightly different > function to calculate the distance. > > distEisen<- function(x, use = "pairwise.complete.obs") { > co.x<- cor(x, use = use) > dist.co.x<- 1 - co.x > return(as.dist(dist.co.x)) > } > > HTH > \Heidi > >> *If* i knew how to implement these in heatmap.2() i would. I assume the >> arguments: "hclustfun =" and "distfun =" could be modified to deliver >> what ever i want, including Pearson Correlation coefficient and linkage >> average linkage methods. But it would be trial, error and googling to >> acheive this. Which is my next step, unless some one has the patience to >> provide an example or point me to one (cheers!). So this is my sole >> reason for using heatplot over heatmap.2 (or heatmap). >> >>> >>> library(gplots) >>> my.mat<- matrix(rnorm(250), nrow=50, ncol=5) >>> heatmap.2(my.mat, col=c(redgreen(256)), dendrogram="col", trace="none", >>> density.info="none") >>> >>> will give you something similar to >>> >>> heatplot(my.mat, dend="both", cols.default=TRUE) >>> >>> only with the colours you want, and without the row dendrogram (but with >>> row reordering). >>> >>> If you want to use heatplot() because you actually have some made4 >>> defined >>> object and not a matrix to begin with, then note that the first couple >>> of >>> lines in the heatplot() code are >>> >>> data<- array2ade4(dataset) >>> data<- as.matrix(data) >>> >>> Presumably you can call something like that on your data first, to get a >>> matrix that can be plotted with heatmap.2 >>> >>> On an different note, then green-black-red isn't really the optimal >>> colour >>> scheme due to the number of red/green colour blind people. Consider >>> using >>> maybe blue-black-yellow or something else instead. >> >> I never considered this aspect. Indeed i will. >> >> Thanks again for your thoughts and any further ones you (or anyone else) >> may have, >> >> Karl >> >> The RColorBrewer >>> package also has a range of pre-defined colour schemes. Check out e.g. >>> display.brewer.all(9) >>> >>> HTH >>> \Heidi >>> >>>> Hi Jim, >>>> >>>> Sorry- guilty of impatience/laziness. >>>> >>>> Find below what i ran on my machine and the errors i received. I >>>> thought >>>> given heatplot() is based on heatmap.2(), i could just use some >>>> heatmap.2() arguments to get what i wanted. I failed and lack the >>>> understanding to modify heatplot() myself to achieve my goal- >>>> red-BLACK-green colors and suppressing drawing of the row dendrogram. >>>> >>>> All tips greatly appreciated, >>>> >>>> Karl >>>> >>>> ###run local### >>>> # Bioc examples heatplot() returned errors: >>>> library(made4) >>>> my.mat<- matrix(rnorm(250), nrow=50, ncol=5) >>>> >>>> # 1. default works fine: >>>> windows() >>>> plot.new() >>>> heatplot(my.mat, dend="both", cols.default=TRUE) >>>> >>>> # 2. attempt red/black/green palette with hetamap.2 col= arg >>>> windows() >>>> plot.new() >>>> heatplot(my.mat, dend="both", cols.default=FALSE, col = >>>> c(redgreen(256))) >>>> >>>> # 3. attempt to supress drawing row dendrogram tree not NOT row >>>> reordering >>>> windows() >>>> plot.new() >>>> heatplot(my.mat, dend="both", cols.default=TRUE, dendrogram = "column") >>>> >>>> ###end >>>> >>>> #2. returned: >>>> >Error in heatmap.2(data, Colv = Colv, Rowv = Rowv, col = plotcols, >>>> scale = scale, : >>>> > formal argument "col" matched by multiple actual arguments >>>> >>>> #3. returned: >>>> >Error in heatmap.2(data, Colv = Colv, Rowv = Rowv, col = plotcols, >>>> scale = scale, : >>>> > formal argument "dendrogram" matched by multiple actual arguments >>>> >>>> > sessionInfo() >>>> R version 2.10.0 (2009-10-26) >>>> i386-pc-mingw32 >>>> >>>> locale: >>>> [1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United >>>> States.1252 >>>> [3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C >>>> >>>> [5] LC_TIME=English_United States.1252 >>>> >>>> attached base packages: >>>> [1] grid grDevices datasets splines graphics stats >>>> tcltk >>>> utils methods >>>> [10] base >>>> >>>> other attached packages: >>>> [1] made4_1.20.0 scatterplot3d_0.3-29 RColorBrewer_1.0-2 >>>> ade4_1.4-14 >>>> [5] gplots_2.7.4 caTools_1.10 bitops_1.0-4.1 >>>> gtools_2.6.1 >>>> [9] reshape_0.8.3 plyr_0.1.9 gdata_2.7.1 >>>> svSocket_0.9-48 >>>> [13] TinnR_1.0.3 R2HTML_1.59-1 Hmisc_3.7-0 >>>> survival_2.35-7 >>>> >>>> loaded via a namespace (and not attached): >>>> [1] cluster_1.12.1 lattice_0.17-26 svMisc_0.9-56 tools_2.10.0 >>>> >>>> >>>> >>>> On 8/6/2010 7:34 PM, James W. MacDonald wrote: >>>>> Hi Karl, >>>>> >>>>> You need to show us the function call you used rather than saying what >>>>> you did. The errors indicate that you used the 'col' argument twice in >>>>> the first attempt and the 'dendrogram' argument twice in the second >>>>> attempt. But without seeing exactly what you did, nobody can say for >>>>> sure. >>>>> >>>>> Best, >>>>> >>>>> Jim >>>>> >>>>> >>>>> >>>>> On 8/6/10 11:55 AM, Karl Brand wrote: >>>>>> Esteemed BioC Users, >>>>>> >>>>>> Could some one help me obtain the garden variety microarray heatmap >>>>>> "red-*black*-green" palette in heatplot()? Using- >>>>>> >>>>>> lowcol="green", highcol="red" >>>>>> >>>>>> -has no black transition between red and green. >>>>>> >>>>>> Since its based on heatmap.2, i thought i could achieve this by >>>>>> passing the following argument to heatplot(): >>>>>> >>>>>> cols.default=FASLE #to suppress default color palette >>>>>> col = c(redgreen(256)) >>>>>> >>>>>> but get the error- >>>>>> >>>>>>> Error in heatmap.2(data, Colv = Colv, Rowv = Rowv, col = plotcols, >>>>>> scale = scale, : >>>>>>> formal argument "col" matched by multiple actual arguments >>>>>> >>>>>> I'd also like to suppress drawing of the row dendrogram (but NOT the >>>>>> reordering or rows), but again the heatmap.2 argument to do this- >>>>>> >>>>>> dendrogram = "column" #draw only column dendrograms >>>>>> >>>>>> returned- >>>>>> >>>>>>> Error in heatmap.2(data, Colv = Colv, Rowv = Rowv, col = plotcols, >>>>>> scale = scale, : >>>>>>> formal argument "dendrogram" matched by multiple actual arguments >>>>>> >>>>>> With thanks in advance, cheers, >>>>>> >>>>>> Karl >>>>>> >>>>> >>>> >>>> -- >>>> Karl Brand >>>> Department of Genetics >>>> Erasmus MC >>>> Dr Molewaterplein 50 >>>> 3015 GE Rotterdam >>>> T +31 (0)10 704 3457 |F +31 (0)10 704 4743 |M +31 (0)642 777 268 >>>> >>>> _______________________________________________ >>>> Bioconductor mailing list >>>> Bioconductor at stat.math.ethz.ch >>>> https://stat.ethz.ch/mailman/listinfo/bioconductor >>>> Search the archives: >>>> http://news.gmane.org/gmane.science.biology.informatics.conductor >>>> >>> >>> >> >> -- >> Karl Brand >> Department of Genetics >> Erasmus MC >> Dr Molewaterplein 50 >> 3015 GE Rotterdam >> T +31 (0)10 704 3457 |F +31 (0)10 704 4743 |M +31 (0)642 777 268 >> > > -- Karl Brand <k.brand at="" erasmusmc.nl=""> Department of Genetics Erasmus MC Dr Molewaterplein 50 3015 GE Rotterdam P +31 (0)10 704 3409 | F +31 (0)10 704 4743 | M +31 (0)642 777 268