-
- Those I am planning to attend/attended:
- 2008-10-21 UKRDS Leeds Focus Group Meeting
- 2008-10-06 European Grid Initiative Meeting, London, UK.
- Other relevant events:
-
- Browsing
- Miscellanea
- Geomorphometrics and Neural Network Classification of Roof Types Research with Sadhvi
- e-Science
- MoSeS
- GENESIS and PhD
- ...coding to produce a model of society...
- PhD Meeting
- MB, AE, AT
- MB suggests the thesis should focus on agent based retail models as this is work I will be asked to do as part of GENESIS
- Some things to consider in retail modelling:
- There are many different types of product.
- How is, where people shop, influenced by things other than residential location?
- Where people work/school/visit effects their retail choice.
- Shopping is a leisure activity often combined with other activities.
- Online shopping is done more by specific groups.
- AE keen to retain a focus on validation
- Modelling issues
- Data
- Do we have data about what is sold where and to who?
- What is the right level of complexity?
- What are the basic atomic agent units?
- What different retail products/collections/activities can be regarded?
- Can the model examine if hypotheses along the lines of the following can be investigate:
- The fewer retail options, the more expensive the goods the more purposeful the trip?
- Other activities influence retailing, which to explicitly include?
- Individuals, households, regions, socio-economic groups?
- Synching with others
- Others in MASS and in the community are developing similar models how do we fit in with this?
- Use of Kirk Harland’s Framework
-
- Browsing
- e-Science
- MoSeS
- Provided latest results for Leeds to Mark:
- Interestingly the total household population in this is 683404 whereas Mark has a result where this is 701219 a difference of 17815:
- There might be many reasons contributing to this difference:
- The maximum Household size in the Household SAR is 12.
- It might be that Household SAR HRPs for specific age groups only have small household sizes whereas in Leeds there are areas made up with many such households (e.g Households of more than 5 in size and with HRP age 70).
- It may be that a spatial map based look at this difference helps identify certain types of areas (e.g. student areas) which might give a clue to why this is the case.
- Checked that the right number of households and the right sized communal establishment populations are in the result.
- The total communal establishment population is 10291.
- Trying to figure out why programs running on NGS are running out of heap space...
-
- Browsing
- e-Science
- MoSeS
- Trying to figure out why programs running on NGS are running out of heap space...
- Completed code refactoring that aimed to tidy the code and check no fundamental differences with jobs run on NGS and those run on PC.
- Spoke to Shiv who recomended a maximum heap size of 1700m on NGS.
- GENESIS and PhD
-
- Browsing
- Miscellanea
- e-Science
- MoSeS
- Trying to figure out why programs running on NGS are running out of heap space...
- -XX:-UseGCOverheadLimit -Xmx4g -Xms4g
- Loading Population 00FY 190 out of 434 java.lang.OutOfMemoryError: Java heap space...
- NeSS
- Informal chat with Rudradeb Mitra about Work Package 4.1
- We have a much clearer understanding of the Work Package status and how to proceed with our collaboration.
- Rudradeb needs to check back with Pete Edwards about a few things before we make any changes to wiki content.
- We are both a bit unlcear as to D4.1.1a, but have a clear tack on D4.1.3.
- Rudradeb/PolicyGrid has useful experience with workflows and is interested in D4.1.5.
- Kepler is being used in PolicyGrid by Eduardo Pignotti.
- In previous work Rudradeb has implemented a workflow engine (based on colored heirarchical petri nets) specially keeping business processes in mind.
- I mentioned that Paul had been working with Taverna.
- Rudradeb hopes to arrange to visit Paul, Junaid and me in Leeds.
- Browsing:
- GENESIS and PhD
-
- Browsing
- Miscellanea
- MASS
- Paper discusiion about:
- O'Sullivan, D., Haklay, M. (2000) Agent-based models and individualism: is the world agent-based?. In Environment and Planning A 32, no. 8: 1409 - 1425, doi:10.1068/a32140.
- e-Science
-
-
- Browsing
- Miscellanea
- Completed university travel to work survey:
- e-Science
- MoSeS
- GENESIS and PhD
- ...coding to produce a model of society...
- ...overcoming limitations of Gregorian and Julian Calendars...
-
- Browsing
- Miscellanea
- Distributed data and grid computing discussion with Nick Malleson
- Nick is using the Derby database to store data and wondered how best to do this for a distributed processing job.
- I have no experience of writing to databases from within MPJ Express environments, so he is a bit on his own investigating this.
- I described the options for data management I have employed including passing all the data in messages within MPJ, writing to local filespace and writing to federated filespace.
- We discussed the pros and cons for various data management and parallel processing strategies.
- We looked at some of the code for and I demonstrated running a job on NGS.
- Nick is going to get a UK e-Science Certificate to be able to run on NGS as well as WRG :-)
- e-Science
-
- Browsing
- Miscellanea
- Geomorphometrics and Neural Network Classification of Roof Types Research
- Discussion with Gordon Mitchell about EPSRC Small Grants Funding...
- e-Science
- MoSeS
- Progress with calculating missing results:
- Trying to get NGS code running to calculating missing results:
- It seemed that all Population loaded, but there was a failure as before but the Error was caught...
- java.lang.OutOfMemoryError: Java heap space
- NeSS
- GENESIS and PhD
-
- Browsing
- Miscellanea
- e-Science
- MoSeS
- Trying to get NGS code running to calculating missing results:
- Java options: -Xmx3000m -Xms3000m
- Different error:
- Loading Population 38UE 209 out of 434
- java.lang.OutOfMemoryError: GC overhead limit exceeded
- Investigating what to do about this:
- GENESIS and PhD
- ...coding to produce a model of society...
-
- Browsing
- Miscellanea
- PC Crashed at around 16:35
- CSAP Meeting
- Presentation on a current ESRC project on projecting ethnic populations for local areas in the UK.
- Phil Rees described the ethnic projection model.
- Pia Wohland described ethnic mortality estimation work.
- Pete Boden described new immigration estimation work.
- Paul Norman described ethnic fertility estimation work.
- This all seemed very relevant to MoSeS demographic modelling work.
- I asked if there estimations would be made available to us and the answer seemed to be yes :-)
- e-Science
- MoSeS
- NeSS
- Feedback on Workpackage 3.1 Metadata Languages – Standards and Formats document version 0.4 section g) GeoSpatial Metadata:
- Three key overlapping organisations for geospatial metadata are:
- The Open Geospatial Consortium Metadata Working Group [1];
- The International Standards Organisation Technical Committee 211 Geographic information/Geomatics [2]; and,
- The World Wide Web Consortium Metadata and Resource Description [3] - in particular the W3C Geospatial Incubator Group [4].
-
"Data quality and provenance are critical to deriving real benefit from geography and geolocation on the Web.
A general approach to descriptive metadata is modeled in ISO 19115
and rules for XML encoding are specified in ISO 19139." [4]
- The Federal Geographic Data Committee (FGDC) of the United States of America provides a profile of ISO 19115 for their National Spatial Data Infrastructure.[5]
- References:
- GENESIS and PhD
- ...coding to produce a model of society...
-
- Browsing
- e-Science
- MoSeS
- Trying to get NGS code running to calculating missing results:
- Java options: -Xmx2000m -Xms2000m
- Error:
- java.lang.OutOfMemoryError: Java heap space
- GENESIS
- ...coding to produce a model of society...
-
- Browsing
- Miscellanea
- Steven Taylor (Steve) now at Natural England asked about automated identification of grips in moorland areas
- e-Science
- MoSeS
- GENESIS
- Started coding to produce a model of society...
-
-
-
- Browsing
- Miscellanea
- e-Science
-
- e-Science
- European Grid Initiative (EGI) Meeting
-
- Browsing
- Miscellanea
- Geomorphometrics and Neural Network Classification of Roof Types Research with Sadhvi
- Helping Kelvin Peh with TRMM data conversion
- TRMM data is available via:
- There is a readme file that details what the data is:
- Kelvin wants to retrieve from the TRMM data monthly total monthly rainfall at the following point location for the period Jan 2007 to Sept 2008:
- Basically what is wanted amounts to 21 estimates of total precipitation.
- Are the available TRMM data for the same time periods?
- Need to study the readme file that details what the data is:
- If not, some further work to compile these estimates might be necessary:
- It is expected that this will be aggregation.
- Example TRMM data file which may contain this point:
- What we want is a utility for converting the data.
- Does a this exist?
- If so is it easy to use like the Lat, Lon conversion tool referenced above?
- Do the components exists to make this?
- If so can someone guide us in how to put these together and estimate how much work this is?
- Investigation to find tools that can handle the TRMM data format:
- Is the TRMM data format more commonly known as HDF Format?
- If so what type(s)/version(s)?
- The HDF Group
- Geotools
- Other potential tools:
- Although Kelvin only wants to do one set of conversions for his study, he reckons such a utility would be generally useful for the group he works with...
-
- Miscellanea
- Geomorphometrics and Neural Network Classification of Roof Types Research with Sadhvi
- Trying to understand how to use some Neural Network code developed by Jianhui Jin
- River Basins Processes and Management Research Cluster Meeting
-
- Browsing
- Miscellanea
- Geomorphometrics and Neural Network Classification of Roof Types Research with Sadhvi
- Error loading data j 43, i 268
- MASS meeting
- Signed up on http://www.academia.edu/