# covariance

##### Covariance and Correlation

`cov()`

and `var()`

form the variance-covariance matrix. `cor()`

forms
the correlation matrix. `cov2cor()`

scales a covariance matrix into a
correlation matrix.

- Keywords
- methods

##### Usage

```
# S4 method for ddmatrix
cov(x, y = NULL, use = "everything",
method = "pearson")
```# S4 method for ddmatrix
var(x, y = NULL, na.rm = FALSE, use)

# S4 method for ddmatrix
cor(x, y = NULL, use = "everything",
method = "pearson")

# S4 method for ddmatrix
cov2cor(V)

##### Arguments

- x, y, V
numeric distributed matrices.

- use
character indicating how missing values should be treated. Acceptable values are the same as

`R`

's, namely "everything", "all.obs", "complete.obs", "na.or.complete", or "pairwise.complete.obs".- method
character argument indicating which method should be used to calculate covariances. Currently only "spearman" is available for

`ddmatrix`

.- na.rm
logical, determines whether or not

`NA`

's should be dealth with.

##### Details

`cov()`

forms the variance-covariance matrix. Only
`method="pearson"`

is implemented at this time.

`var()`

is a shallow wrapper for `cov()`

in the case of a
distributed matrix.

`cov2cor()`

scales a covariance matrix into a correlation matrix.

##### Value

Returns a distributed matrix.

##### Examples

```
# NOT RUN {
spmd.code = "
library(pbdDMAT, quiet = TRUE)
init.grid()
x <- ddmatrix('rnorm', nrow=3, ncol=3), bldim=2
cv <- cov(x)
cv
finalize()
"
pbdMPI::execmpi(spmd.code = spmd.code, nranks = 2L)
# }
```

*Documentation reproduced from package pbdDMAT, version 0.5-1, License: GPL (>= 2)*