Difference between revisions of "Simple R example"
From Gcube Wiki
(→Processing rational) |
|||
Line 5: | Line 5: | ||
The process takes one OGC WFS getFeature request passed as reference and use R to generate a plot as shown below: | The process takes one OGC WFS getFeature request passed as reference and use R to generate a plot as shown below: | ||
− | [[File: | + | [[File:FRA_TUNA_YEAR_5deg.png]] |
− | The OGC WFS server is hosted by IRD, the GetCapabilities URL is here: [http://mdst-macroes.ird.fr:8080/constellation/WS/wfs/TestNono?REQUEST=GetCapabilities&SERVICE=WFS&VERSION=1.1.0www.example.com | + | The OGC WFS server is hosted by IRD, the GetCapabilities URL is here: [http://mdst-macroes.ird.fr:8080/constellation/WS/wfs/TestNono?REQUEST=GetCapabilities&SERVICE=WFS&VERSION=1.1.0www.example.com IRD GetCapabilities] |
== Processing == | == Processing == | ||
− | The WPS process uses WPS-hadoop in pseudo-cluster mode. | + | === Resources === |
+ | |||
+ | The WPS process uses WPS-hadoop in pseudo-cluster mode. It fruits the hadoop streaming approach where the WFS URLs are pipped into a bash script that: | ||
+ | - retrieves the WFS response via a curl GET request and stores the output (GML file). | ||
+ | - This GML file is passed as argument to the R executable script (provided by IRD). | ||
+ | - The output is published on the HDFS storage and returned to the client as an HTTP reference to the result | ||
+ | |||
+ | === Interface === | ||
+ | |||
+ | The DescribeProcess URL is available here: | ||
+ | |||
+ | === GET invocation === | ||
+ | |||
+ | The GET invocation URL is: | ||
+ | |||
+ | === POST invocation === | ||
+ | |||
+ | The POST invocation is provided by: |
Revision as of 17:55, 27 November 2012
Contents
Simple R example
Processing rational
The process takes one OGC WFS getFeature request passed as reference and use R to generate a plot as shown below:
The OGC WFS server is hosted by IRD, the GetCapabilities URL is here: IRD GetCapabilities
Processing
Resources
The WPS process uses WPS-hadoop in pseudo-cluster mode. It fruits the hadoop streaming approach where the WFS URLs are pipped into a bash script that: - retrieves the WFS response via a curl GET request and stores the output (GML file). - This GML file is passed as argument to the R executable script (provided by IRD). - The output is published on the HDFS storage and returned to the client as an HTTP reference to the result
Interface
The DescribeProcess URL is available here:
GET invocation
The GET invocation URL is:
POST invocation
The POST invocation is provided by: