Example with weather, dates and trust - Redigeerimiste ajalugu

Tanel – 14. november 2016, kell 13:54

2016-11-14T13:54:51Z

Tanel – 14. november 2016, kell 13:52

2016-11-14T13:52:17Z

Tanel: Uus lehekülg: '

2016-11-14T13:37:31Z

Uus lehekülg: '<pre> </pre>'

Uus lehekülg

@@ 1. rida: / 1. rida: @@
 <pre>
 clear(auto).
 set(prolog_style_variables).
 set(hyper_res).
@@ 96. rida: / 49. rida: @@
 % rdfdq is a rdf with date and qualifier:
-% rdfdq(subject,property,object,date,qualifier)
+% rdfdq(subject,property,object,qualifier,date)
 -rdfd(X,"temperature",Y,U) | -$FLT(Y,"-0") | rdfdq(X,"weatherfeature","badness",1,U).

←Vanem redaktsioon		Redaktsioon: 14. november 2016, kell 13:52
1. rida:		1. rida:
	<pre>		<pre>
		+	clear(auto).
		+	set(prolog_style_variables).
		+	set(hyper_res).
		+	set(factor).
		+	clear(print_given).
		+	set(print_kept).
		+	assign(stats_level, 0).
		+	assign(max_seconds, 10).
		+
		+	list(sos).
		+
		+	% temperature in Tallinn is -5 at day one with trust 1:
		+	% here "id1" is an id of the encoded fact
		+
		+	rdf("id1","subject","Tallinn").
		+	rdf("id1","property","temperature").
		+	rdf("id1","object","-1").
		+	rdf("id1","date",1). % 1 is today
		+	rdf("id1","trust","1"). % 1 is full trust
		+
		+	% combine the reified triplets to a fact encoding
		+	% also date and trust where
		+	% rdfdt(subject,property,object,date,trust)
		+
		+	-rdf(X,"subject",Y) \| -rdf(X,"property",Z)
		+	\| -rdf(X,"object",V) \| -rdf(X,"date",W) \|
		+	-rdf(X,"trust",O) \|
		+	rdfdt(Y,Z,V,W,O).
		+
		+	% temperature the next day is the same as today,
		+	% but with twice smaller trust: also, do not derive
		+	% anything with trust less than 0.2
		+
		+	-rdfdt(X,"temperature",Y,U,T) \| -$FLT("0.2",T) \| rdfdt(X,"temperature",Y,$SUM(U,1),$FPROD(T,"0.5")).
		+
		+	% rdfd is rdf with date but no trust:
		+
		+	rdfd("Tallinn","temperature","-5",2). % tomorrow
		+
		+	rdfd("Tallinn","weatherfeature","stormy",2). % tomorrow
		+
		+	% if stormy, then also windy:
		+
		+	-rdfd(X,"weatherfeature","stormy",U) \| rdfd(X,"weatherfeature","windy",U).
		+
		+	% various degrees of badness:
		+	%
		+	%clear(auto).
		+	set(prolog_style_variables).
		+	set(hyper_res).
		+	set(factor).
		+	clear(print_given).
		+	set(print_kept).
		+	assign(stats_level, 0).
		+	assign(max_seconds, 10).
		+
		+	list(sos).
		+
		+	% temperature in Tallinn is -5 at day one with trust 1:
		+	% here "id1" is an id of the encoded fact
		+
		+	rdf("id1","subject","Tallinn").
		+	rdf("id1","property","temperature").
		+	rdf("id1","object","-1").
		+	rdf("id1","date",1). % 1 is today
		+	rdf("id1","trust","1"). % 1 is full trust
		+
		+	% combine the reified triplets to a fact encoding
		+	% also date and trust where
		+	% rdfdt(subject,property,object,date,trust)
		+
		+	-rdf(X,"subject",Y) \| -rdf(X,"property",Z)
		+	\| -rdf(X,"object",V) \| -rdf(X,"date",W) \|
		+	-rdf(X,"trust",O) \|
		+	rdfdt(Y,Z,V,W,O).
		+
		+	% temperature the next day is the same as today,
		+	% but with twice smaller trust: also, do not derive
		+	% anything with trust less than 0.2
		+
		+	-rdfdt(X,"temperature",Y,U,T) \| -$FLT("0.2",T) \| rdfdt(X,"temperature",Y,$SUM(U,1),$FPROD(T,"0.5")).
		+
		+	% rdfd is rdf with date but no trust:
		+	% rdfdq(subject,property,object,date)
		+
		+	rdfd("Tallinn","temperature","-5",2). % tomorrow
		+
		+	rdfd("Tallinn","weatherfeature","stormy",2). % tomorrow
		+
		+	% if stormy, then also windy:
		+
		+	-rdfd(X,"weatherfeature","stormy",U) \| rdfd(X,"weatherfeature","windy",U).
		+
		+	% various degrees of badness:
		+
		+	% rdfdq is a rdf with date and qualifier:
		+	% rdfdq(subject,property,object,date,qualifier)
		+
		+	-rdfd(X,"temperature",Y,U) \| -$FLT(Y,"-0") \| rdfdq(X,"weatherfeature","badness",1,U).
		+
		+	-rdfd(X,"temperature",Y,U) \| -$FLT(Y,"-20") \| rdfdq(X,"weatherfeature","badness",2,U).
		+
		+	-rdfd(X,"temperature",Y,U) \| -$FLT(Y,"0") \| -rdfd(X,"weatherfeature","windy",U) \| rdfdq(X,"weatherfeature","badness",3,U).
		+
		+	end_of_list.

	</pre>		</pre>

Example with weather, dates and trust - Redigeerimiste ajalugu

Tanel – 14. november 2016, kell 13:54

Tanel – 14. november 2016, kell 13:52

Tanel: Uus lehekülg: ' '

Tanel: Uus lehekülg: '

'