[R-sig-Geo] Error ENFA: Error in eigen(Se) : infinite or missing values in 'x'

[Clement Calenge]

Dear Leandro,

On 02/11/2013 02:06 PM, Leandro Abade wrote:
> Dear all,
>
> I am trying to run an ENFA using ADE package though I am stuck in this
> error, and I do not understand what could be the problem.
>
> Error in eigen(Se) : infinite or missing values in 'x'
>
> I get the error message after the following step:
>
>> dataenfa1 <- data2enfa(var,locs)
>> (enfa1 <- enfa(pc, dataenfa1$pr, + scannf = FALSE))
> Error in eigen(Se) : infinite or missing values in 'x'
>
> Any thoughts or help would be much appreciated.

This happens when the covariance matrix associated with the used 
resource units (pixels) is singular, i.e. it cannot be inverted. This 
happens when one of the columns of the table Z – containing the value of 
the variables (columns) in the used resource units (rows) – can be 
predicted by a linear combination of the other columns.

Several problems in the data can lead to this error message, but the 
most likely is that you have in your data one habitat type that is not 
used at all by the species. In this case, one of the columns of Z is 
populated only by zeros, and the associated covariance matrix cannot be 
inverted. Actually, if this habitat type is not used, you cannot measure 
the habitat suitability from the data. Therefore, you have to consider 
that your focus species does not use this environmental type, and to 
delete it from the dataset (var in your example).
HTH,

Clément Calenge

-- 
Clément CALENGE
Cellule d'appui à l'analyse de données
Direction des Etudes et de la Recherche
Office national de la chasse et de la faune sauvage
Saint Benoist - 78610 Auffargis
tel. (33) 01.30.46.54.14