1 / 12

Towards an Algebra of Grid Functions

Towards an Algebra of Grid Functions. Bill Howe. Claims. All our data can be modeled using grids (discretizations of a domain) and functions over the discrete elements All our data products can be specified as manipulations of these grid functions bind subgrid regrid.

eamon
Télécharger la présentation

Towards an Algebra of Grid Functions

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Towards an Algebra of Grid Functions Bill Howe

  2. Claims • All our data can be modeled using • grids (discretizations of a domain) and • functions over the discrete elements • All our data products can be specified as manipulations of these grid functions • bind • subgrid • regrid

  3. Vocabulary: Grids • A k-dimensional cell, or k-cell can be thought of as just a polygon • 0-cell : ‘vertex’ = c • 1-cell : ‘edge’ = (c1, c2) • d-cell: ‘element’ (or just ‘cell’) = (c1, c2,...,cm) • AgridG is a sequence of sets of k-cells [C0,C1,...] • k-cell(G) denotes the (k+1)th set in the sequence, Ck. • dim(G) is the highest k for which k-cell(G) . • G1 G2is the cross product of G1 and G2 2-cell 1-cell

  4. Vocabulary: Grid Functions • GFk a : ( G, f ), where • G is a grid, • f : k-cell(G)  a where a is some type. (scalar float, vector of int, etc.) • grid(GFk a) = G • dom(GFk a) = kcell(G)

  5. Examples of f geometry : 0-cell(G)  (x,y,z) salinity : 2-cell(G)  float neighbors(i, j) : i-cell(G)  {j-cell(G)} measure(i) : i-cell(GF0(x,y,z))  float //length, area, etc. column : k-cell(G1)  {k-cell(G1 G2)}

  6. Bind bind( f )[GFk a] = GFk (a,b) where f : k-cell(G)  b b (b,x) a Bind(f) c (c,y) (a,w) d (d,z) gfZ = bind(vgrid, depths) gfXY = bind(hgrid, geometry) gfBath = bind(gfXY, bath)

  7. Subgrid subgrid(p)[GFk a] = GFk a where p : a  Bool gfXY = bind(hgrid, nodes) gfZ = bind(vgrid, depths) gfBath = bind(gfXY, bath) crossGrid = gfXYb  gfZ corieGrid = subgrid(d > b?, crossGrid)

  8. Regrid Three components: • target grid, G • cell mapping, g where g : k-cell(G)  { j-cell( grid( GFk b ) ) } • aggregate function, f where f : {a}  b Regrid( G, g, f )[GFja] = GFkb

  9. target grid = hgrid cell mapping = g : e e vgrid where e 2-cell(GF) aggregate = max Example: max salinity gfSalt = bind(salt)[hgrid  vgrid] gfMaxSalt = regrid(hgrid, g, max)[gfSalt]

  10. Example: salinity gradient salinity • target grid = same • cell mapping = halo • aggregate function = gradient gfAreas = bind(measure(2))[hgrid  vgrid] gfGeo = bind(geometry)[gfAreas] gfSalt = bind(salinity)[gfGeo] gfSaltGrad = regrid(G, halo, gradient)[gfSalt]

  11. Example: plume volume • target grid = unit_grid • cell mapping = g where g : e {3-cell(G)} • aggregate = sum float gfElev = bind(elev)[corieGrid] gfOcean = subgrid(ocean?)[gfElev] gfPlume = subgrid(salt<26)[gfOcean] gfVol = bind(measure(3))[gfPlume] gfPlumVol = regrid(unit, g, sum)[gfVol]

  12. Grid Functions • Grid functions can model • Spatial data sets • Image data • Plots • Timeseries • Independent of representations • Arrays • Objects • Relations • File Formats (CORIE, NetCDF, etc.)

More Related