correlation plot in color

palette <- setNames(object = c(?red?, ?green3?, ?black?, ?blue?), nm =levels(u$Treat))

Hi Manuel,


I am glad that you managed to plot the correlation plots. Then, as for 
the colour argument, bg is the argument for setting the background 
color, not the symbols color; the argument controlling the symbols 
colour in plots is color. The error you get seems to be due to the 
fact that your command is missing? a square bracket at the end.


Still, I have never used the PerformanceAnalytics package before and I 
just do not know how to make the chart.Correlation function to use the 
color argument. My guess is that it is not designed to take such argument.


Salvador


Salvador S?NCHEZ-COL?N

An independent consultant

Statistics, GIS, RS



En Dom, 22 Marzo, 2020 en 16:42, Manuel Esteban Lucas Borja 
<ManuelEsteban.Lucas at uclm.es> escribi?:
Para: Salvador SANCHEZ COLON
Cc: bfalevlist at gmail.com; r-sig-ecology at r-project.org

Thanks Salvador,

It Works properly and I was able to do the correlation plots excluding 
the categorical variable as you mentioned. In relation to colours I 
found the following information from other websites:

data(iris)

pairs(iris[1:4], main = "Data(iris) -- 3 Especies", pch = 21,bg =

c("red", "green3", "blue")[codes(iris$Species)])

So the argument bg should be used for adding colours to the 
correlation plot.

bg= c(?red?, ?green3?,?blue?)[codes(iris$Species)

I tried this with my dataset:

chart.Correlation(Libro[,1:4], histogram = T, pch = 19, bg= c(?red?, 
?green3?, ?black? ?blue?)[codes(Libro$Treat))

Results:

Error: unexpected input in "chart.Correlation(u[,1:4], histogram = T, 
pch = 19, bg= c(?"

                                                          
?blue?)[codes(u$Treat))

Error: unexpected input in " ?"

Thanks a lot for your time!

Best

Manuel

*De:*Salvador SANCHEZ COLON <salvadorsanchezcolon at prodigy.net.mx>
*Enviado el:* domingo, 22 de marzo de 2020 19:25
*Para:* Manuel Esteban Lucas Borja <ManuelEsteban.Lucas at uclm.es>
*CC:* bfalevlist at gmail.com; r-sig-ecology at r-project.org
*Asunto:* Re: [R-sig-eco] correlation plot in color

Hola Manuel:

As you have loaded the ggplot2 package, one simple way to do the 
correlation plots (one at the time though) is, for example for your 
pair of variables EMF and your first MSD axis:

*ggplot*|(|data =|Libro) |+||
*geom_point*|(|mapping =||*aes*|(|x =|EMF, |y =|MSD1, |color 
=|Treat))|

I hope this helps.

Salvador

An independent consultant

Biostatistics, GIS, RS

En Dom, 22 Marzo, 2020 en 10:34, Manuel Esteban Lucas Borja 
<ManuelEsteban.Lucas at uclm.es <mailto:ManuelEsteban.Lucas at uclm.es>>
escribi?:

Para: Bede-Fazekas ?kos; r-sig-ecology at r-project.org 
<mailto:r-sig-ecology at r-project.org>

Dear ?kos Bede-Fazekas
Sorry for the missed information. Please see the packages I used:

library(readxl)
library(dplyr)
library(ggplot2)
library(GGally)
library(Hmisc)
library(corrplot)
library(PerformanceAnalytics)

Then I import the attached dataset (Libro1) containing 5 variables: 
EMF, bio, MSD1, MSD2 (all of them numerical) and Treat (categorical).

Then I wrote:

chart.Correlation(Libro1, histogram = F, pch = 19)

And got the error:

Error in pairs.default(x, gap = 0, lower.panel = panel.smooth, 
upper.panel = panel.cor) :
non-numeric argument to 'pairs'

And I would like to obtain the plot you can see on the doc file, where 
correlation points are divided by a category (red, gren and blue)

Thanks a lot for your help
Manuel



-----Mensaje original-----
De: R-sig-ecology <r-sig-ecology-bounces at r-project.org
<mailto:r-sig-ecology-bounces at r-project.org>> En nombre de 
Bede-Fazekas ?kos Enviado el: domingo, 22 de marzo de 2020 16:57
Para: r-sig-ecology at r-project.org <mailto:r-sig-ecology at r-project.org>
Asunto: Re: [R-sig-eco] correlation plot in color

Dear Manuel,

since neither a reproducible example or your dataset are provided, nor 
the details of your problem (results of sessionInfo(); the package of 
corrMatOrder(); the error message you got) are known by us, it is a 
bit hard to help you. Anyway, I guess, that your data is not in 
categorical
(factor) but in ordinal (ordered) scale. If you convert your ordoered 
data to numeric with as.numeric(), then Spearman and Kendall rank 
correlations can be used.

HTH,
?kos Bede-Fazekas
Hungarian Academy of Sciences


2020.03.22. 14:29 keltez?ssel, Manuel Esteban Lucas Borja ?rta:

Dear Drew,
Thank you very much for your help. The thing is that when I import

the dataset containing categorical and numeric variables and then I 
want to calculate the correlation graphic, Rstudio says that it is not 
posible as categorical variables can not be used for correlation. So, 
which is the code for calculating correlation plots including a 
categorical variable, which will be them used for adding colours 
depending of the levels of this categorical factor?

Thanks a lot for your advise, I really apreciate your help.
Best
Manuel




Manuel Esteban Lucas Borja
Profesor Contratado Doctor Interino / Assistant Professor 
Coordinador de Relaciones Internacionales del Campus de 
Albacete/Coordinator for International Relations on Albacete Campus

Universidad de Castilla-La Mancha
E.T.S. Ingenieros Agr?nomos y de Montes / Higher Technical School of 
Agricultural and Forestry Engineering | Campus Universitario s/n | 
C.P. 02071 Albacete
Tfno: 967 599 200 | Ext: 2818 |

<mailto:ManuelEsteban.Lucas at uclm.es>>;

r-sig-ecology at r-project.org <mailto:r-sig-ecology at r-project.org>

<r-sig-ecology at r-project.org <mailto:r-sig-ecology at r-project.org>>

Subject: Re: [R-sig-eco] correlation plot in color

Manuel,

T may be getting interpreted as TRUE, not your variable name. Try

renaming the variable to something else and see if that helps.

--
Drew Tyre

He/him/his

While I may be sending this email outside my normal office hours, I 
have no expectation to receive a reply outside of yours.

School of Natural Resources
University of Nebraska-Lincoln
416 Hardin Hall, East Campus
3310 Holdrege Street
Lincoln, NE 68583-0974

phone: +1 402 472 4054
fax: +1 402 472 2946
email: atyre2 at unl.edu <mailto:atyre2 at unl.edu> 
http://snr.unl.edu/tyre http://drewtyre.rbind.io

?Anything worth doing is worth doing badly.?

* G. K. Chesterton

?On 3/21/20, 8:32 PM, "R-sig-ecology on behalf of Manuel Esteban

Lucas Borja" <r-sig-ecology-bounces at r-project.org on behalf of 
ManuelEsteban.Lucas at uclm.es 
<mailto:r-sig-ecology-bounces at r-project.org on behalf of 
ManuelEsteban.Lucas at uclm.es>> wrote:

Dear all,



First of all, I hope this email will find you well.



I am working with a biological dataset (please see in attachment)

composed by 5 variables. The columns EMF, Biodiversity, MDS1 and MDS2 
are numerical variables whereas T is a categorical one.

I would like to obtain a correlation chart showing correlation

coefficients, histograms and different colours for each observation 
following the categorical variable "T". Please see below the initial 
chart I got but in which observations are not classified in different 
colours following the categorical variable "T".

This is the code I used with any positive result back.



corrMatOrder (dataset, histogram = T, pch = 19, method= spearman, 
family="CM Roman", col=c("red", "green3", "blue",
"black")[code.levels(dataset$T)])



I work with Rstudio. Thank you very much for your time and help.

Best



Manuel