Pearson's correlation matrix with p-values in R

#Computes Pearson's correlation matrix with p-values in R
library(metan)
All numeric variables
View(data_ge2)
ALL=corr_coef(data_ge2)
ALL
plot(ALL)
Select variables
Some=corr_coef(data_ge2, EP, EL, CD, CL,KW,ED,EL)
Some
plot(Some)
############

[@wakjiratesfahun3682]

This is an additional tutorial on the topic of correlation. Please click on the provided link to access it. th-cam.com/video/rlNzizM-rck/w-d-xo.html

[@somy30]

Thank you so much! This is the best correlation matrix with P-values in R. It is not only easy to generate, but also reader-friendly.

[@wakjiratesfahun3682]

Thanks for the positive response.

[@magazinefariasevoce-aminha6326]

Thank you very much! Very useful and easy to do it. Congratulations !! You helped me a lot ! 😁😁

[@mebratuify]

Thank you, our brother. W also expect the Spearman correlation analysis and drawing graphs. Thank you so much.

[@wakjiratesfahun3682]

It is not difficult to do. Anyways I will publish.

[@andreasvoldstad4516]

So useful! Any way to determine the ordering of the elements in the correlation matrix?

[@wakjiratesfahun3682]

Okay ,I will publish a video within two hours.

[@kariiamba7324]

Thankyou very much. Precise and informative.

[@camillep3925]

Thank you very much for this very useful tutorial !!
Is there anyway we can add the adjusted p-value to the correlogram instead of the p-value ?
Thank you ! :)

[@syuenlim1627]

Thanks Dr Tesfahun, Im just wondering if this function could be used for my biological variables that was analysed with PCA ?

[@wakjiratesfahun3682]

yes you can.

[@mesfinhailemariam9684]

Very useful as usual!!!

[@wakjiratesfahun3682]

Thanks Mesfene

[@abrehamfeyisa3606]

Thanks a lots Dr Tesfahun, really I have get a great concept of R software manipulations skills & knowledge from the video you have been post & share for us! Keep it up please!!! Saying this I would like to know how I will be accessed & have download a softy copy of data (data_ge2) used on example of your video?; i.e, which already attached under the package of 'metan' by the author Tiago Olivoto.
thank you, with regards!

[@wakjiratesfahun3682]

write.csv(data_ge2,"C:/Users/AAA/Desktop//Abreham.csv")

[@abrehamfeyisa3606]

@@wakjiratesfahun3682 Thanks for your quick response Dr. I was write this Syntax and submit to software but it will say that "object data_ge2" not found. May be to clarify my question I will find to download the soft copy of personal data 'data_ge2' which was used as an example on the R software package by the author. I need it only for using of other example & exercise may be in case I need. thanks

[@wakjiratesfahun3682]

@@abrehamfeyisa3606 I think it is important. Let's do a tutorial on it. I will release a video within an hour.

[@mikiallen7733]

nice effort sir , nonetheless if I want to create more or less the same analysis but dynamically , in other words , using on-line type of calculation , while updating such correlations on the fly , what elements/extra tools and concepts within R ecosystem then one shall consider in making that happen ? your input is highly appreciated

[@deryaakpinar5412]

Thanks for the helpful instruction. I have to determine the correlation between two data sets so that the number of values in each column is not the same. I tried to input "NA" for the non-valued column, but the code did not work. Do you have a solution such as this dataset, to create the correlation coefficient matrix? Thanks in advance for your answer!

[@wakjiratesfahun3682]

The classic correlation coefficient is defined for paired observations. It is not defined for unpaired observations. If you have two variables with different sizes, they are not paired, and it is not possible to calculate the correlation coefficient of both variables.

[@deryaakpinar5412]

@@wakjiratesfahun3682 I searched right after I commented, and as you mentioned, it has to be an equal sample size to calculate the correlation coefficient. I appreciate the quick response, though!

[@wakjiratesfahun3682]

@@deryaakpinar5412 Thank you.

[@praharshinikumar1429]

Thanks for the video! I wanted to ask if there is a way to only compute p-values since I already have a correlation matrix?

[@wakjiratesfahun3682]

Please look this link. th-cam.com/video/rlNzizM-rck/w-d-xo.html

[@praharshinikumar1429]

Thanks :) Also, what kind of statistical analysis would you recommend for comparing three sets of correlation matrices (each set is 22x22 matrix, and they comprise of fisher transformed correlation coefficients@@wakjiratesfahun3682

[@habtamuhailu9900]

guddaa galatoomaa waan gaarii irraa baradhee jira

[@wakjiratesfahun3682]

Ayee!

[@ibokopaul6385]

Thank you but please is it possible to order the variables on both X and Y axis when using corr_coef? eg. let X start with EP and end in ED, and Y starts from EP and end in ED just like in X axis

[@wakjiratesfahun3682]

I just published a video . Check my recent video.

[@Aesthetics10000]

Hey! Can you please guide about how to make mix correlation plot (heatmap) along with scatterplot and histogram?

[@tigerproject]

How can I remove boxes with non-significant p values please? Thanks!

[@xxs010]

Thank man!!!

[@wakjiratesfahun3682]

🥂

[@nilimeshmridha7616]

Thanks for the video. Can you share how to increase font size in the matrix plot?

[@adityapratapsingh6068]

Can we keep the sequence of variables in the heatmap according to our own choice? Like in your case EP, EL, CD etc. While working with more number of variables, there sequence always gets jumbled up. Any remedy for this?

[@wakjiratesfahun3682]

Yes, it is possible in R. However, before doing so, you need to arrange your data according to your preferences. Once the data is properly organized, you can instruct R to correlate the variables in the same manner as the data.
M=cor(data)
corrplot(M)
corrplot(M,order ='original' )

[@ibokopaul6385]

@@wakjiratesfahun3682 The problem with this is that it does not give the same information as corr_coef map. It does the arrangement properly but fails to present it in the same format as in corr_coef. ie., it does not include the information on p values, coeffifients etc

[@wakjiratesfahun3682]

I'm not sure why this occurred for you. When I reviewed mine, it appeared to be accurate. Adding p-values to the graph can be quite cluttered, and I believe the level of significance is adequate.

[@user-ln6iq5wr4q]

Hi,
Thank you for the usefull video. Does anyone know how to make the order pf variables on the graph the same as the order of variables in the table ?
Thanks

[@wakjiratesfahun3682]

Yeah, you have to specify order='original'

[@wakjiratesfahun3682]

corrplot(M
type="upper",
order="original" ,
tl.cex = 0.5,
col=colorRampPalette(c("blue4", "white", "firebrick1"))(100),
p.mat = pM$P,
sig.level = 0.05,
insig = "blank")

[@ruchithruchi5009]

can i change the order of the combination in the plot according to my convenience? if yes, can you help me out. thankyou

[@wakjiratesfahun3682]

Please look this link. th-cam.com/video/rlNzizM-rck/w-d-xo.html

[@sedjrobienvenukpeki2991]

Thanks a lots

[@wakjiratesfahun3682]

Thanks!

[@monisham4277]

Sir, i want a character to be present at last even though i have arranged data as such in plot it is appearing in the middle
how to do that

[@wakjiratesfahun3682]

You are asking me about how to re arrange

[@Plantscience786]

How I can get the visulization in circle form instead of numbers with p values.

[@wakjiratesfahun3682]

I will share the codes

[@vikrant549]

@wakjiratesfahun3682 Why my color bar is reversed and how to correct it? Thanks.

[@wakjiratesfahun3682]

What type of color you want ?

[@rifathossain9272]

How can I change the color and method to a circular one?

[@wakjiratesfahun3682]

Please find the following link. It answers your question.
th-cam.com/video/zcQbxA1Gmts/w-d-xo.html

[@tesfayemidega4887]

what You are writing is not Visible to understand R script you used.

[@wakjiratesfahun3682]

Thank you for your comment. I will check it.

[@rdatacode]

corrplot, lares packages, ggstatsplot , all can generate correlation matrix

[@wakjiratesfahun3682]

Sure.