[R] Memory errors when using QCA package

[Adrian Dusa]

Allan Engelhardt <allane <at> cybaea.com> writes:
> 
> It is a little stupid, but the length of a vector is limited to 2^31-1 
> entries on any platform.  A matrix is stored as a vector, so the product 
> of all dimensions is also limited to 2^31-1.
> 
> Allan.
> 
> Matthew Gwynne wrote:
> > Hi,
> >
> > I have been using the QCA package, in particular the "eqmcc" function
> > and I am having some issues when trying to use this to minimise a
> > particular boolean function.
> >
> > The  boolean function in question has 16 variables, and I am providing
> > the full truth table for the function (65536 with 256 true entries),
> > in the following way :\
> > [...snip...]

Indeed, thanks very much Allan.
There is a question though for Matthew: why do you try to provide the full truth
table? The function eqmcc() proves its strength exactly for the cases when most
of the causal combinations are unknown (as I believe it is your case).
I would suggest to run the analysis using only the observed combinations, and
eqmcc() will know what to do next.

QCA accepts both truth tables and observed combinations as entry data, but (as
you can see), there are limitations for extemely large truth tables. Should you
need further assistance, I'd be glad to help.

I hope this helps,
Adrian