-
- MoSeS
- Set Beowulf running to produce UK dataset... Hoping for a complete result in the new year!
-
-
-
-
- Larry Kerschberg, Hanjo Jeong and Wooju Kim (2006) Emergent Semantics in Knowledge Sifter: An Evolutionary Search Agent based on Semantic Web Services
- GRADE Scoping a Geospatial Repository for Academic Deposit and Extraction
- Lots of overlap and crossover in these newsletters:
- Started using GeoVISTA Studio
- It seems that Security from OWS-4 can be used for much of what is required for GEESE part 2 and the moves in OWS-5 look good for JISC OGC Grid Collision.
- I don't understand yet how Grid Security is to be incorporated.
- I understand how Shibboleth can be useful and eScience Proxy Certificate tokens, but is there any software from the Grid community working with XACML or SAML or something that can do the DRM?
- Perhaps the benefits of Grid Security is all about securing/encrypting the data so it can be processed in a safe virtual environment on computational resources that are systems administered in say China for example.
- Read SEcurE access to GEOspatial services Draft Project Plan
- Read Outline Workplan Proposal in response to the JISC Grid Enabling National Data Centres Initiative
- Ian Turton's Community MapBuilder WMS client JSR168 portlet
- Again offered to send a pointer/the code.
- Ian reckonned I should hassle him more.
- It should be open source and available from the GeoVista repository, but as it's not finished it probably is not.
- Ian has moved on to other things, but is receptive to the idea of us building on this :-)
- CeWolf built on top of JFreeChart is what Ian and Alistair used for VoB
- ConceptVISTA - GeoVista teams ontology creation and visualization tool
- Got to grips with Blogging from outside university thanks to Mike Crabtree :)
- Updated OGC TC meeting notes
- Travel to OGC TC meeting
- Flights over the West coast of Scotland, Greenland and over lake Michigan into Chicogo, then on from Chicago at dusk across the mountains and into San Diego.
- EEE Seminar
- Bryan Lawrence's Blog
- Geomorphometrics
- Nidderdale process was producing odd results as the data were odd:
- The data were at 10 metre resolution, but all values the same for blocks of 5X5 cells.
- Aggregated data produced much more sensible results and because the data 25 times smaller the results are there for the entire study area.
- Contacted Kiarash to register for SOSoRNET
- Geomorphometrics
- Nidderdale process was going OK, but slow so restarted using a larger chunksize (up from 64 to 256 rows and columns).
- Consider submiting abstract for Hydrogeomorphology BSG Conference - Joe Holden reckons attendance can be funded with some special SoG RAE fund.
- MoSeS
- SOSoRNET
- Geomorphometrics
- Nidderdale process was going OK, but slow so restarted using a larger chunksize (up from 64 to 256 rows and columns).
- MoSeS
- Spotted two errors in my code for Handling Table CAS002 :-(
- aggregation not correct for Ages 55to59
- This will have effected results at aggregate levels...
- Code fixed, but need to re-aggregate data and produce a new set of results
- Data for females single age 16 to 19 loaded as females single age 20 to 24.
- All results are effected :-(
- Lack of data checking that we didn't do is coming back to haunt us...
- In our rush to get the algorithms working we did not check out data IO was robust.
results
- Meeting with Mark
- My persuasion pays off and Mark agrees that I need a new PC :)
- I am to ask Mike Crabtree to spec me one up for around £1.5K.
- A new PC will allow me to work faster, run Netbeans 5.5 with my projects and use its profiler software to help optimise the MoSeS Population Initialisation program. The savings could be huge...
- I outlined overlap between OGC Grid Collision, NCeSS e-Infrastructure and MoSeS and described SEE-GEO and SAW-GEO projects.
- Portal
- Paul working on contectual mapping
- I encouraged Mark to ask Paul to use EDINA provided data rather than scrape anything from online services like local site, Google or Yahoo Maps.
- I am to focus on producing more comparable results with Marks for a meeting on Thursday
- Age of HRP (should be straight lines as constrained)
-
- Mark interested in comparing OA data aggregated to Ward with Ward data from Census, and plans to set me to this later...
- I explained that we currently do the comparison on an ad hoc basis.
- What might be better is if we choose a census table and produce the aggregated results that can be compared directly with that census table.
- Mark interested in comparing OA data aggregated to Ward with Ward data from Census, and plans to set me to this later...
- I explained a generic aggregation for single variables from the SARs using a (key,value) mapping.