on 01-11-2016 8:35 AM
Hi
How can I create a simple correlation matrix in SAP PA? Pearson's R, Cramer's V and Spearman's Rank Correlation coeffeicient would be needed
Cheers,
Uwe
Hello Uwe,
This collection of Custom R Components includes a Correlation Matrix.
http://scn.sap.com/docs/DOC-62119
You can download the component as .spar file. The article contains details on how to import it.
Many Greetings
Andreas
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hello Uwe, is your question answered?
Best regards,
Antoine
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Uwe,
In addition to Andreas' suggestion below, in case you are working on a PA + HANA architecture, you could use the Multivariate Statistics PAL within HANA.
This function calculates several basic multivariate statistics including covariance matrix and Pearson’s correlation coefficient matrix. The function treats each column as a data sample.
Here is an extract from the PAL documentation for your reference:
Covariance Matrix
The covariance between two data samples (random variables) x and y is:
Suppose that each column represents a data sample (random variable), the covariance matrix Σ is defined as the covariance between any two random variables:
where X=[X1,X2,...,Xn].
Pearson’s Correlation Coefficient Matrix
The Pearson’s correlation coefficient between two data samples (random variables) X and Y is the covariance of Xand Y divided by the product of standard deviation of X and the standard deviation Y:
Similar to the covariance matrix, the Pearson’s correlation coefficient matrix Σ is:
where X=[X1,X2,...,Xn]
Regards,
Surya
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
81 | |
10 | |
10 | |
9 | |
7 | |
6 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.