Since I slightly modified the cacheBounds initialization when translating nodes when adding the features wanted by Marco, I regenerate the mondrian health report.
Slightly improvement from the one I sent at the beginning of the month. I discovered an opportunity to gain many % when rendering nodes. I will work on that soon.
Report produced on 2009-12-28T13:32:14+00:00 Benchmark ManyNode (simple rendering of nodes) : 100 nodes => 16 ms 200 nodes => 54 ms 300 nodes => 108 ms 400 nodes => 187 ms 500 nodes => 301 ms 600 nodes => 446 ms 700 nodes => 574 ms 800 nodes => 755 ms 900 nodes => 949 ms 1000 nodes => 1148 ms 1600 nodes => 3043 ms Benchmark ManyEdges (simple rendering of edges) : 10 edges => 1 ms 20 edges => 5 ms 30 edges => 13 ms 40 edges => 180 ms 50 edges => 50 ms 60 edges => 79 ms 70 edges => 117 ms 80 edges => 174 ms 90 edges => 256 ms 100 edges => 342 ms 200 edges => 4432 ms 300 edges => 35438 ms 55.08 % of methods are covered Progress from last time: -0.96 %
Cheers, Alexandre
Nice :)
Doru
On 28 Dec 2009, at 13:36, Alexandre Bergel wrote:
Since I slightly modified the cacheBounds initialization when translating nodes when adding the features wanted by Marco, I regenerate the mondrian health report.
Slightly improvement from the one I sent at the beginning of the month. I discovered an opportunity to gain many % when rendering nodes. I will work on that soon.
Report produced on 2009-12-28T13:32:14+00:00 Benchmark ManyNode (simple rendering of nodes) : 100 nodes => 16 ms 200 nodes => 54 ms 300 nodes => 108 ms 400 nodes => 187 ms 500 nodes => 301 ms 600 nodes => 446 ms 700 nodes => 574 ms 800 nodes => 755 ms 900 nodes => 949 ms 1000 nodes => 1148 ms 1600 nodes => 3043 ms Benchmark ManyEdges (simple rendering of edges) : 10 edges => 1 ms 20 edges => 5 ms 30 edges => 13 ms 40 edges => 180 ms 50 edges => 50 ms 60 edges => 79 ms 70 edges => 117 ms 80 edges => 174 ms 90 edges => 256 ms 100 edges => 342 ms 200 edges => 4432 ms 300 edges => 35438 ms 55.08 % of methods are covered Progress from last time: -0.96 %
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Yesterday is a fact. Tomorrow is a possibility. Today is a challenge."
On Dec 28, 2009, at 1:36 PM, Alexandre Bergel wrote:
Since I slightly modified the cacheBounds initialization when translating nodes when adding the features wanted by Marco, I regenerate the mondrian health report.
Slightly improvement from the one I sent at the beginning of the month. I discovered an opportunity to gain many % when rendering nodes.
what is the idea? in general
I will work on that soon.
Report produced on 2009-12-28T13:32:14+00:00 Benchmark ManyNode (simple rendering of nodes) : 100 nodes => 16 ms 200 nodes => 54 ms 300 nodes => 108 ms 400 nodes => 187 ms 500 nodes => 301 ms 600 nodes => 446 ms 700 nodes => 574 ms 800 nodes => 755 ms 900 nodes => 949 ms 1000 nodes => 1148 ms 1600 nodes => 3043 ms Benchmark ManyEdges (simple rendering of edges) : 10 edges => 1 ms 20 edges => 5 ms 30 edges => 13 ms 40 edges => 180 ms 50 edges => 50 ms 60 edges => 79 ms 70 edges => 117 ms 80 edges => 174 ms 90 edges => 256 ms 100 edges => 342 ms 200 edges => 4432 ms 300 edges => 35438 ms 55.08 % of methods are covered Progress from last time: -0.96 %
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
I discovered an opportunity to gain many % when rendering nodes.
what is the idea? in general
Each time you translate a node, you need to update the size of the canvas. Updating the canvas takes a bit of time (O(n^2)). When you do a layout, each node is translated by the algorithm. The idea is to update the canvas after having done the complete layout.
Cheers, Alexandre
On Dec 28, 2009, at 2:52 PM, Alexandre Bergel wrote:
I discovered an opportunity to gain many % when rendering nodes.
what is the idea? in general
Each time you translate a node, you need to update the size of the canvas. Updating the canvas takes a bit of time (O(n^2)). When you do a layout, each node is translated by the algorithm. The idea is to update the canvas after having done the complete layout.
I see. What you call the size of the canvas is the entity containing all the entity? The top level entity?
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Canvas is the morph place on which the graphs are drawn. Root is the node that contains all other nodes and edges.
Cheers, Doru
On 28 Dec 2009, at 15:52, Stéphane Ducasse wrote:
On Dec 28, 2009, at 2:52 PM, Alexandre Bergel wrote:
I discovered an opportunity to gain many % when rendering nodes.
what is the idea? in general
Each time you translate a node, you need to update the size of the canvas. Updating the canvas takes a bit of time (O(n^2)). When you do a layout, each node is translated by the algorithm. The idea is to update the canvas after having done the complete layout.
I see. What you call the size of the canvas is the entity containing all the entity? The top level entity?
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Some battles are better lost than fought."
-
Alexandre Bergel -
Stéphane Ducasse -
Tudor Girba