[R] Spatstat: xy binary data into mask type to use in owin(mask=)

[Adrian Baddeley]

If 'X' is a data frame containing columns 'x', 'y' and 'value, try

         m <- with(X, tapply(value, list(y,x), all))
         z <- im(m, xcol=sort(unique(X$x)), yrow=sort(unique(X$y)))
         w <- as.owin(z)

This will only work if the x, y values form a rectangular grid in some 
order.

Adrian Baddeley

=============================================
Javier PB wrote:
>
>> Dear users,
>>
>> I am trying to export polygons from Arcmap into Spatstat to run some
>> simulations using functions available in Spatstat package.
>>
>> One particular area to be exported is formed by a number of polygons
>> defining the external boundaries of the area (as a groups of islands) 
>> and a
>> number of polygons inside the previous ones,  as “holes” not to be
>> considered as part of the area.
>>
>> I have managed to export one of these areas using owin(poly=list()),
>> including in the list the number of external boundaries and holes, 
>> but the
>> number of polygons that constitute each area is large and I was 
>> wondering if
>> this could be easily done using owin(mask=my_area), using only one 
>> single
>> matrix file "my_area" with TRUE, FALSE values defining the whole area.
>>
>> Here is where I got stuck. The file I have created in ArcMap contains 
>> three
>> columns: x and y coordinates, and the values at points x[i], y[j] (1 
>> if the
>> area is TRUE and 0 is the area is FALSE).
>>
>> How can I covert this xy file into a mask type "my_area" that I can read
>> using owin(mask=my_area)?