Prolog: nat(s(s(0))) to "normal" decimal numbers

Question

If an integers is defined like this in Prolog:

nat(0).
nat(s(X)) :- nat(X).

How do I convert such a number to a decimal number?

The input is for example:

s(s(s(0)))

I probably should add that I am very new to Prolog.

EDIT: I tried it this way:

nat(0).
nat(s(X)) :- nat(X).

convert(N, C) :-
    C is C + 1,
    nat(N),
    convert(N, C).

What I want is a "normal" numerical representation like 0, 1, 2, 3, ... in order to use it as an index or to be able to use compares. — Willi, Jun 16 '20 at 20:37
Prolog is not C, **C = C +1** always fails in Prolog because of unification. You must use another name **C1 is C + 1** for example — joel76, Jun 16 '20 at 21:12
See also: https://stackoverflow.com/questions/8954435/convert-peano-number-sn-to-integer-in-prolog — David Tonhofer, Jun 16 '20 at 21:56

score 1 · Answer 1 · answered Jun 16 '20 at 20:38

1

There are basically two cases you need to cover:

The base case thus looks like:

convert(0, 0).

the recursive case thus looks like:

convert(S(X), N) :-
    convert(X, N1),
    ….

where … is thus a step you need to do to determine N given N1. I leave this as an exercise.

answered Jun 16 '20 at 20:38

Willem Van Onsem

Thanks, I am just going to throw out this recursive definition of numbers in the entire code base since I can't properly calculate with it. – Willi Jun 16 '20 at 21:09
It would be possible to write enough functions to calculate with them - that's the whole point of this model for basic mathematics - but it would be neither convenient nor efficient. – aschepler Jun 16 '20 at 22:19

1 Answers1