Closed AlfredWinter closed 2 years ago
There are only 190 node pairs with overlapping edges now:
select count(*)
{
select distinct ?s ?o
from <http://www.snik.eu/ontology>
{
?s ?p ?o.
?s ?q ?o.
?s a owl:Class.
?o a owl:Class.
filter(?p!=?q).
filter(isIRI(?o)).
filter(?p!=meta:subTopClass&&(?q!=meta:subTopClass))
}
}
From 10228 node pairs with edges:
select count(*)
{
select distinct ?s ?o
from <http://www.snik.eu/ontology>
{
?s ?p ?o.
?s a owl:Class.
?o a owl:Class.
}
}
So less than 2% of edges should have this problem now.
Update: There are also 75 cases with reversed directions:
select count(*)
{
select distinct ?s ?o
from <http://www.snik.eu/ontology>
{
?s ?p ?o.
?o ?q ?s.
?s a owl:Class.
?o a owl:Class.
filter(?p!=?q).
filter(isIRI(?o)).
filter(?p!=meta:subTopClass&&(?q!=meta:subTopClass))
}
}
text-margin-x shifts along the x axis of the canvas and is thus not suitable. source-text-margin-x does not seem to shift anything, we probably don't have a source label.
As it seems to be a nontrivial time investment to fix this, wait first whether Franziska Jahns Meta Model property hierarchy will eliminate most of those cases.
Alle Paare von Knoten, bei denen sich Kanten in Hin- und Rückrichtung überlappen:
Um die Tabelle voll zu sehen, bitte im Firefox https://userstyles.org/styles/97661/wide-my-github aktivieren
select distinct ?s ?o (GROUP_CONCAT(DISTINCT ?p; SEPARATOR=", ") AS ?hin) (GROUP_CONCAT(DISTINCT ?q; SEPARATOR=", ") AS ?zurueck)
from <http://www.snik.eu/ontology>
{
?s ?p ?o.
?o ?q ?s.
filter(?s<?o).
?s a owl:Class.
?o a owl:Class.
filter(?p!=?q).
filter(isIRI(?o)).
filter(?p!=meta:subTopClass&&(?q!=meta:subTopClass))
} group by ?s ?o
Alle Paare von Knoten, die sich in der gleichen Richtung überlappen:
select distinct ?s ?o (GROUP_CONCAT(DISTINCT ?p; SEPARATOR=", ") AS ?p) from http://www.snik.eu/ontology { ?s ?p ?o. ?s ?q ?o. filter(?s<?o). ?s a owl:Class. ?o a owl:Class. filter(?p!=?q). filter(isIRI(?o)). filter(?p!=meta:subTopClass&&(?q!=meta:subTopClass)) } group by ?s ?o
Moved to the "next" milestone as we are waiting for @FranziskaJahn to create a property hierarchy of the meta model.
The property hierarchy does not reduce this further. However it doesn't seem to be worth the effort in relation to the small number of cases now so I move this to later.
The edge "bb:ServerCluster meta:entityTypeComponent bb:Server" is incorrect.
Details Hier überlagern sich zwei Bezeichnungen für die Relation und man kann es nicht mehr lesen. Siehe auch #59 .AW