new feature list 2 sp

Just a quick bit of code to improve context.learn(). We can now learn a rule that is a list.

Here is the python:
def list_2_sp(one):
  r = superposition()
  if type(one) == list:
    for x in one:                                # what do we want to do if type(x) is not int, float or string?
      if type(x) == int or type(x) == float:
        r += ket("number: " + str(x))
      elif type(x) == str:
        r += ket(x)
  return r
And a quick demonstration of it in action:
# define some example lists:
list1 = [2,3,5,7,11,13]
list2 = ["cat","dog","horse","rat","mouse","lion","horse"]
list3 = ["a","b",37,2.1828,"a","fish","a",37]

# test learn code:
context.learn("list-of","small primes",list1)
context.learn("list-of","common animals",list2)
context.learn("list-of","test elements",list3)

# see what we have learnt:
context.print_universe()
Outputs (NB: the repeated elements with coeff > 1):
list-of |small primes> => |number: 2> + |number: 3> + |number: 5> + |number: 7> + |number: 11> + |number: 13>
list-of |common animals> => |cat> + |dog> + 2|horse> + |rat> + |mouse> + |lion>
list-of |test elements> => 3|a> + |b> + 2|number: 37> + |number: 2.1828> + |fish>
So, simple enough. Just makes using the python context.learn() a little cleaner. eg, instead of:
context.learn("friends","Fred",ket("Sam") + ket("Frank") + ket("Jim") + ket("Rob"))
we can now do:
context.learn("friends","Fred",["Sam","Frank","Jim","Rob"])
Of course, if you need your kets to have coeff other than 1, then you need to do it the old way, with the explicit ket(string,value). Though I suspect most of the time we can get away with just using lists.


Home
previous: new operators times by divide by plus minus
next: new operators mod k and is mod k

updated: 19/12/2016
by Garry Morrison
email: garry -at- semantic-db.org