Reading a csv file of city distance in prolog?

Question

I have a file like this as csv

Now i want this file to create facts such as this in prolog as

distance(A,E,3)
distance(A,G,5)
and so on.

As per the other answers I have tried csv_read_file("test.csv", Rows, [functor(citydist), arity(5)]), maplist(assert, Rows), print(Rows).

but this is returning just a row wise data. How to read it as an adjacency graph?

Answer 1

Assuming the CSV looks something like :

"Distance","E","G","C","B"
"A","3","5","7","3"
"B","4","5","6","0"
"C","4","1","0","6"
"D","5","8","3","6"

:- [library(csv)] .
:- [library(lists)] .

%! distance(CITY_A,CITY_B,DISTANCE)
%
% dynamic facts asserted based upon csv file .

:- dynamic distance/3 .

:- op(1,'xfy','csv_') .


%! csv(FILE0)
%
% read the CSV file `FILE0` and add facts to the database .

csv(FILE0)
:-
(start) csv_ (FILE0)
.

(start) csv_ (FILE0)
:-
csv:csv_read_file(FILE0,[HEADER|ROWss]) ,
row__to__list(HEADER,HEADERs) ,
(loop) csv_ (HEADERs,ROWss) 
.

(loop) csv_ (_HEADERs,[]) 
:-
true
.

(loop) csv_ (HEADERs,[ROW|ROWss]) 
:-
row__to__list(ROW,ROWs) ,
lists:nth1(1,ROWs,CITY_A) ,
QUERY_A=(lists:nth1(NTH,ROWs,DISTANCE)) ,
QUERY_B=(NTH > 1) ,
QUERY_C=(lists:nth1(NTH,HEADERs,CITY_B)) ,
QUERY=(QUERY_A,QUERY_B,QUERY_C) ,
ASSERT=assertz(distance(CITY_A,CITY_B,DISTANCE)) ,
forall(QUERY,ASSERT) ,
(loop) csv_ (HEADERs,ROWss) 
.

row__to__list(ROW,ROWs)
:-
ROW=..[_|ROWs]
.

/*
?- csv('data/distances.csv').
true ;
false.

?- listing(distance).

distance('A', 'E', 3).
distance('A', 'G', 5).
distance('A', 'C', 7).
distance('A', 'B', 3).
distance('B', 'E', 4).
distance('B', 'G', 5).
distance('B', 'C', 6).
distance('B', 'B', 0).
distance('C', 'E', 4).
distance('C', 'G', 1).
distance('C', 'C', 0).
distance('C', 'B', 6).
distance('D', 'E', 5).
distance('D', 'G', 8).
distance('D', 'C', 3).
distance('D', 'B', 6).

?-
*/

Answer 2

Reminder: Capital valued entities are logic variables, so your facts should have lower case letters a, b, c, etc.

Supposing you know what are your nodes, once the distances have been entered into Prolog, here's how to find the adjacent nodes for a given Node:

adjacencies(Node, Adjacencies) :-
    findall(Adj, distance(Node, Adj, _), Adjacencies).

So with some distance facts like in your illustration,

?- adjacencies(a, L).
L = [e,g,c,b]

Does this anwser your question ?