CELLMAPS : Summary page

 

Jacques Paris  May 2003

 

 

You can view the support PDF document in French only directly form here, or download it from the left panel.

Your contribution is vital for the sound development of this project by providing it with solid and well documented foundations and widening the range of future applications. Your input can feed two streams: general references and specific topics. Specific items could become the basis for a “Guide” in that area of “cellmaps with MI” while references could contribute to various chapters of this guide.

 

Besides, a special place must be made for MI tools specifications, i.e. MBX applications or MI compatible software that would be useful in cellmap production and management; nothing forbids that some of those tools could be also useful beyond that specialized perspective (e.g. re-classification tool).

 

I will receive your input and “dispatch” it among the relevant “headings”. I will try to keep a synthesis page in both languages in order to make the collected information as widely available to MI users as possible. But I would not able to offer you a valid product without YOUR involvement that, I know, will be to constrained by your resources and available time but that is crucial in that collective enterprise.

 

The following scheme is an initial suggestion. It will certainly be revised as your contributions will help refining the issues. It will keep serving as “Table of Contents” for a number of pages that will be added and updated as time goes by.

 

 

A short definition of a cellmap

 

A cellmap is formed exclusively of basic elements (regions) covering a study area that is not necessarily rectangular; these elements have a constant shape (e.g. square, hexagonal) but may not always be of the same area. The table attached to the map can contain a number of variables, each filled with values related to the basic cells.

 

It is somewhere between raster maps (rectangular coverage of even size squared pixels with a single variable by image) and mosaic maps (obtained by intersecting several region maps into irregular shapes and with variables extracted from each original maps)

 

Cellmaps are often used to collect information in a predetermined structure in order to facilitate uniform analyses across the study area.

 

As the most common cellmap applications are rectangular arrays of equal area square cells, the term of grid is often used instead. This term is not the most adequate because it carries an image that is too restrictive.

 

 

1 – General references

 

1 – 1 Cellmaps projects

 

Example should include the following points :

 

Coverage : study area, basic grid

Data: types and mode of acquisition

Analyses : desired products and processing procedures

 

It should also include references (sites, publications, attached documents) commentaries and critique ... , all that can be relevant to this project.

 

1 – 2 Theoretical references

 

Any document that deals with procedures, rules or models that could be of use n this domain.

 

Identification of the document or of the source, table of relevant contents, availability and/or accessibility

 

1 – 3 Software of interest

 

Any software allowing for cellmap processing or offering processing procedures of similar nature.

 

Identification of the software, details of relevant features, availability and/or accessibility.

 

Possibilities for interfacing with MI if some cellmap functionalities are available; potential constraints.

 

1 – 4 Tools

 

Any application working within  MI and capable of tasks that are relevant in the cellmap perspective.

 

Identification of the, description of relevant functionalities, availability and/or accessibility

 

 

 

2 – Elements for a practical guide

 

2 – 1 Cellmap foundations

 

Questions dealing with the conception of a specific cellmap :

 

2-1-a  Cell shape

 

2-1-b  Cell size in relation with the data, the phenomena, the expected products.

 

2-1-c  Study area

 

2 – 2 Tabular data of a cellmap

 

Questions relative to data acquisition and value assignment.

 

2-2-a  Direct transfer (ex. : de raster à cellules) with possible merging of pixels (aggregation rules)

 

2-2-b  Procedures of extraction from digital maps : original data attached to punctual, linear or closed. Extraction rules (centroid,..) and value assignment.

 

2-2-c  Data re-coding. Principles, rules and tools of re-classification of original data when first acquired.

 

2 – 3 Pre-processing transformations

 

Data transformations prior to analysis

 

2-3-a  Transformations of single variables (one variable at the). Principles, rules and tools of re-classification of tabular data. (extension of 2-2-c)

 

2-3-b  Generation of complex variables (involving several variables). Principles, rules and tools of generation of complex variables.

 

2 – 4 Analysis

 

2-4-a  Analytical models : aggregative (merging various variables), combinatory (two variables at the time, along a binary tree), multivariate...

 

2-4-b  Rules for systematic processing of all possible situations (e.g. involving binary, categorical and numeric variables)

 

2-4-c  Tools for model implementation.

 

2 – 5 The results

 

Anything related to the presentation of the analysis results.

 

2-5-a  Products as thematic maps, complex layouts and reports

 

2-5-b  Tools supporting the production of the results.

 

 

 

3 – Tool specifications

 

Tools of any form that can be used within MI and that could play a role in cellmaps. It is premature to enter in the details yet; here are some general indications.

 

Functionalities to consider

 

Constraints on inputs (a cell map contains only « region » cells and can be processed sequentially but cells cannot be addressed my row/column addresses)

 

Constraints on products (how results will fit in the final cellmap)