Topic's layer distribution
| Line 5: | Line 5: | ||
For example, if we have a topic with layer distribution vector '''[ 1:1:0 ]''', we can conclude the topic is a merged one and contains topics in first and second layer but the third layer is intact. However, the vector doesn't express the cause of the merge. Topics in Wandora do merge when they share base name, subject locator or subject identifier. | For example, if we have a topic with layer distribution vector '''[ 1:1:0 ]''', we can conclude the topic is a merged one and contains topics in first and second layer but the third layer is intact. However, the vector doesn't express the cause of the merge. Topics in Wandora do merge when they share base name, subject locator or subject identifier. | ||
| − | If we have a layer distibution vector '''[ 1:2 ]''', we can conclude the topic is a merged one but also that the topic in first layer causes two different topics to merge in second layer. This is possible for example when the topic in the first layer shares a base name and a subject identifier with | + | If we have a layer distibution vector '''[ 1:2 ]''', we can conclude the topic is a merged one but also that the topic in first layer causes two different topics to merge in second layer. This is possible for example when the topic in the first layer shares a base name and a subject identifier with two distinct topics in layer 2. |
Some vectors are impossible. For example situation where single layer contains two or more merged topics can never occur as the virtual merge collapses into a real one as soon as topics share identifier. '''[ 2:0:0 ]''' is an example of such impossible layer distribution. | Some vectors are impossible. For example situation where single layer contains two or more merged topics can never occur as the virtual merge collapses into a real one as soon as topics share identifier. '''[ 2:0:0 ]''' is an example of such impossible layer distribution. | ||
Revision as of 12:49, 13 March 2008
Layer distribution vector is an ordered set of numbers representing topic's distribution between layers. Layer distribution vector is used to give a rapid outlook of the topic and how the topic is merged between layers.
Layer distribution vector is shown at the footer, near bottom right corner of the Wandora window. Layer distribution vector contains one number for each layer. First number represents the first layer, second number the second layer etc. If the number in the vector is 0 (zero) current topic doesn't exist in equivalent layer. If number in the vector is 1 (one) current topic has one merged topic in the equivalent layer. If number is 2 topic has two merged topics in the equivalent layer etc. Layer distribution vector contains always positive integer numbers including zero.
For example, if we have a topic with layer distribution vector [ 1:1:0 ], we can conclude the topic is a merged one and contains topics in first and second layer but the third layer is intact. However, the vector doesn't express the cause of the merge. Topics in Wandora do merge when they share base name, subject locator or subject identifier.
If we have a layer distibution vector [ 1:2 ], we can conclude the topic is a merged one but also that the topic in first layer causes two different topics to merge in second layer. This is possible for example when the topic in the first layer shares a base name and a subject identifier with two distinct topics in layer 2.
Some vectors are impossible. For example situation where single layer contains two or more merged topics can never occur as the virtual merge collapses into a real one as soon as topics share identifier. [ 2:0:0 ] is an example of such impossible layer distribution.
Math with Layer distribution
This is very interesting question. Can one calcute anything reasonable with the layer distribution vector? For example has dot product any useful meaning here? We are open for your suggestions and welcome all ideas.
