MultiEOF {CSTools} | R Documentation |

## EOF analysis of multiple variables starting from an array (reduced version)

### Description

This function performs EOF analysis over multiple variables,
accepting in input an array with a dimension `"var"`

for each variable to
analyse. Based on Singular Value Decomposition. For each field the EOFs are
computed and the corresponding PCs are standardized (unit variance, zero mean);
the minimum number of principal components needed to reach the user-defined
variance is retained. The function weights the input data for the latitude
cosine square root.

### Usage

```
MultiEOF(
data,
lon,
lat,
dates,
time = NULL,
lon_dim = "lon",
lat_dim = "lat",
time_dim = "ftime",
sdate_dim = "sdate",
var_dim = "var",
neof_max = 40,
neof_composed = 5,
minvar = 0.6,
lon_lim = NULL,
lat_lim = NULL,
ncores = NULL
)
```

### Arguments

`data` |
A multidimensional array with dimension |

`lon` |
Vector of longitudes. |

`lat` |
Vector of latitudes. |

`dates` |
Vector or matrix of dates in POSIXct format. |

`time` |
Deprecated parameter, it has been substituted by 'dates'. It will be removed in the next release. |

`lon_dim` |
A character string indicating the name of the longitudinal dimension. By default, it is set to 'lon'. |

`lat_dim` |
A character string indicating the name of the latitudinal dimension. By default, it is set to 'lat'. |

`time_dim` |
A character string indicating the name of the temporal dimension. By default, it is set to 'time'. |

`sdate_dim` |
A character string indicating the name of the start date dimension. By default, it is set to 'sdate'. |

`var_dim` |
A character string indicating the name of the variable dimension. By default, it is set to 'var'. |

`neof_max` |
Maximum number of single eofs considered in the first decomposition. |

`neof_composed` |
Number of composed eofs to return in output. |

`minvar` |
Minimum variance fraction to be explained in first decomposition. |

`lon_lim` |
Vector with longitudinal range limits for the calculation for all input variables. |

`lat_lim` |
Vector with latitudinal range limits for the calculation for all input variables. |

`ncores` |
An integer indicating the number of cores to use for parallel computation. The default value is NULL. |

### Value

A list containing:

`coeff` |
An array of principal components with dimensions 'time_dim', 'sdate_dim', number of eof, rest of the dimensions of 'data' except 'lon_dim' and 'lat_dim'. |

`variance` |
An array of explained variances with dimensions 'eof' and the rest of the dimensions of 'data' except 'time_dim', 'sdate_dim', 'lon_dim' and 'lat_dim'. |

`eof_pattern` |
An array of EOF patterns obtained by regression with dimensions: 'eof' and the rest of the dimensions of 'data' except 'time_dim' and 'sdate_dim'. |

`mask` |
An array of the mask with dimensions ('lon_dim', 'lat_dim', rest of the dimensions of 'data' except 'time_dim'). It is made from 'data', 1 for the positions that 'data' has value and NA for the positions that 'data' has NA. It is used to replace NAs with 0s for EOF calculation and mask the result with NAs again after the calculation. |

`coordinates` |
Longitudinal and latitudinal coordinates vectors. |

### Author(s)

Jost von Hardenberg - ISAC-CNR, j.vonhardenberg@isac.cnr.it

Paolo Davini - ISAC-CNR, p.davini@isac.cnr.it

### Examples

```
exp <- array(runif(1280)*280, dim = c(dataset = 2, member = 2, sdate = 3,
ftime = 3, lat = 4, lon = 4, var = 1))
lon <- seq(0, 3)
lat <- seq(47, 44)
dates <- c("2000-11-01", "2000-12-01", "2001-01-01", "2001-11-01",
"2001-12-01", "2002-01-01", "2002-11-01", "2002-12-01", "2003-01-01")
Dates <- as.POSIXct(dates, format = "%Y-%m-%d")
dim(Dates) <- c(ftime = 3, sdate = 3)
cal <- MultiEOF(data = exp, lon = lon, lat = lat, dates = Dates)
```

*CSTools*version 5.2.0 Index]