This operator computes correlation between two contineous vaiables, or between a continuous variable and a categorical factor.
| Input | . |
|---|---|
x-axis |
value of one group of data |
y-axis |
value of other group of data |
| Settings | . |
|---|---|
method |
Correlation test method to be used. |
| Output | . |
|---|---|
cor_coef |
numeric, correlation coefficient calculated per cell |
p_value |
numeric, p-value calculated per cell |
A pearson correlation is performed per cell where the two groups of data to be correlated are found on the x-axis and y-axis.
This operator is ideally suited as an operator for multiple pairwise projections in Tercen.
If the x-axis is non-numeric the operator attempts to convert to a dummy variable (0,1 .., n-1) for a corresponding factor with n levels. This may be most meaningful for a binary factor (n = 2)
See the base::cor.test function of the R package for the documentation.