Traverse query question

[Thomas Müller]

Hello,

for testing i generated a category hierachy ("V_Category") which has 4 levels and 4 categories per level, starting from one ROOT-category (#17:0). The hierachy is build up with "E_CH" category hierachy edges ( parentCategory.addEdge(EEdges.E_CH.name(), childCategory); ). At all 1364 categories.

I also generated 500 documents which have assigned randomly 1 to 10 of the above categories:

 

       nextInt = RandomUtils.nextInt(9) + 1;
        for (int j = 0; j < nextInt; j++) {
            OrientVertex orientVertex = nextRandomCategory().getOrientVertex();
            document.addEdge(EEdges.E_Category.name(), orientVertex);
        }

i want to select the number of documents assigned to a category. What i tried ist:

select count(in('E_Category')) from (traverse out('E_CH'), in('E_Category') from #17:0)

What i expect to get is 500 since i have 500 documets. What i get is 1866 which is 1364(number of categories) +500(number of documents) +2 (?). Even if i change the number of categories this formula seems to be constant: number of categories+number of documents+2. The good thing: the duplicates (one document may have several categories) already seem to be removed (no distinct necessary).

How do i have to change the querry to get only the number of documents which have assigned one or more categories from the given root category and all child categories.

[Luca Garulli]

Hi Thomas,

Could you describe your graph better? 

Lvc@

On 8 March 2015 at 16:22, Thomas Müller <oberin...@googlemail.com> wrote:

Hello,

for testing i generated a category hierachy ("V_Category") which has 4 levels and 5 categories per level, starting from one ROOT-category (#17:0). The hierachy is build up with "E_CH" category hierachy edges ( parentCategory.addEdge(EEdges.E_CH.name(), childCategory); ). At all 1364 categories.

I also generated 500 documents which have assigned randomly 1 to 10 of the above categories:

 

       nextInt = RandomUtils.nextInt(9) + 1;
        for (int j = 0; j < nextInt; j++) {
            OrientVertex orientVertex = nextRandomCategory().getOrientVertex();
            document.addEdge(EEdges.E_Category.name(), orientVertex);
        }

i want to select the number of documents assigned to a category. What i tried ist:

select count(in('E_Category')) from (traverse out('E_CH'), in('E_Category') from #17:0)

What i expect to get is 500 since i have 500 documets. What i get is 1866 which is 1364(number of categories) +500(number of documents) +2 (?). Even if i change the number of categories this formula seems to be constant: number of categories+number of documents+2. The good thing: the duplicates (one cocument may have several categories) already seem to be removed (no distinct necessary).

How do i have to change the querry to get only the number of documents which have assigned one or more categories from the given root category and all child categories.

--

---
You received this message because you are subscribed to the Google Groups "OrientDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email to orient-databa...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Thomas Müller]

Hi Luca,

Thanks for your reply!

My category graph looks like that:

Level 1:

(#17:0_ROOT_Categorie)-['E_CH']->(#17:1_1.Category)
(#17:0_ROOT_Categorie)-['E_CH']->(#17:2_2.Category)
(#17:0_ROOT_Categorie)-['E_CH']->(#17:3_3.Category)
(#17:0_ROOT_Categorie)-['E_CH']->(#17:4_4.Category)

...
Level 2:

(#17:1_1.Category)-['E_CH']->(#17:11_1.1.Category)
(#17:1_1.Category)-['E_CH']->(#17:12_1.2.Category)
(#17:1_1.Category)-['E_CH']->(#17:13_1.3.Category)
(#17:1_1.Category)-['E_CH']->(#17:14_1.4.Category)

...
Level 3:

(#17:11_1.1.Category)-['E_CH']->(#17:111_1.1.1.Category)
(#17:11_1.1.Category)-['E_CH']->(#17:112_1.1.2.Category)
(#17:11_1.1.Category)-['E_CH']->(#17:113_1.1.3.Category)
(#17:11_1.1.Category)-['E_CH']->(#17:114_1.1.4.Category)

... and so on... 4 levels deep and 4 categories per level

I have assigned one to 10 categories to each of my 500 documents:

(#23:1_1.Doc)-['E_Category']->(#17:111_1.1.1.Category)
(#23:1_1.Doc)-['E_Category']->(#17:2_2.Category)
(#23:1_1.Doc)-['E_Category']->(#17:113_1.1.3.Category)

...

(#23:50_50.Doc)-['E_Category']->(#17:4_4.Category)
(#23:50_50.Doc)-['E_Category']->(#17:11_1.1.Category)

...

(#23:500_500.Doc)-['E_Category']->(#17:4_4.Category)
(#23:500_500.Doc)-['E_Category']->(#17:112_1.1.2.Category)
(#23:500_500.Doc)-['E_Category']->(#17:11_1.1.Category)
(#23:500_500.Doc)-['E_Category']->(#17:114_1.1.4.Category)

... and so on, 500 documents.

I want to select the number of documents which have assigned the given category and all child categories of the given category without duplicates. So with given ROOT category (#17:0) i expect to get 500 documents as result with my query but i got 1364 as mentioned above.