` `

Grid gridding

Inverse distance weighted

Description

Inverse distance grid interpolation from irregular distributed points.

Parameters

Points [vector: point]
<put parameter description here>
Attribute [tablefield: any]
<put parameter description here>
Target Grid [selection]

<put parameter description here>

Options:

  • 0 — [0] user defined

Default: 0

Distance Weighting [selection]

<put parameter description here>

Options:

  • 0 — [0] inverse distance to a power
  • 1 — [1] linearly decreasing within search radius
  • 2 — [2] exponential weighting scheme
  • 3 — [3] gaussian weighting scheme

Default: 0

Inverse Distance Power [number]

<put parameter description here>

Default: 2

Exponential and Gaussian Weighting Bandwidth [number]

<put parameter description here>

Default: 1

Search Range [selection]

<put parameter description here>

Options:

  • 0 — [0] search radius (local)
  • 1 — [1] no search radius (global)

Default: 0

Search Radius [number]

<put parameter description here>

Default: 100.0

Search Mode [selection]

<put parameter description here>

Options:

  • 0 — [0] all directions
  • 1 — [1] quadrants

Default: 0

Number of Points [selection]

<put parameter description here>

Options:

  • 0 — [0] maximum number of points
  • 1 — [1] all points

Default: 0

Maximum Number of Points [number]

<put parameter description here>

Default: 10

Output extent [extent]

<put parameter description here>

Default: 0,1,0,1

Cellsize [number]

<put parameter description here>

Default: 100.0

Outputs

Grid [raster]
<put output description here>

Console usage

processing.runalg('saga:inversedistanceweighted', shapes, field, target, weighting, power, bandwidth, range, radius, mode, points, npoints, output_extent, user_size, user_grid)

See also

Kernel density estimation

Description

<put algorithm description here>

Parameters

Points [vector: point]
<put parameter description here>
Weight [tablefield: any]
<put parameter description here>
Radius [number]

<put parameter description here>

Default: 10

Kernel [selection]

<put parameter description here>

Options:

  • 0 — [0] quartic kernel
  • 1 — [1] gaussian kernel

Default: 0

Target Grid [selection]

<put parameter description here>

Options:

  • 0 — [0] user defined

Default: 0

Output extent [extent]

<put parameter description here>

Default: 0,1,0,1

Cellsize [number]

<put parameter description here>

Default: 100.0

Outputs

Grid [raster]
<put output description here>

Console usage

processing.runalg('saga:kerneldensityestimation', points, population, radius, kernel, target, output_extent, user_size, user_grid)

See also

Modifed quadratic shepard

Description

<put algorithm description here>

Parameters

Points [vector: point]
<put parameter description here>
Attribute [tablefield: any]
<put parameter description here>
Target Grid [selection]

<put parameter description here>

Options:

  • 0 — [0] user defined

Default: 0

Quadratic Neighbors [number]

<put parameter description here>

Default: 13

Weighting Neighbors [number]

<put parameter description here>

Default: 19

Left [number]

<put parameter description here>

Default: 0.0

Right [number]

<put parameter description here>

Default: 0.0

Bottom [number]

<put parameter description here>

Default: 0.0

Top [number]

<put parameter description here>

Default: 0.0

Cellsize [number]

<put parameter description here>

Default: 100.0

Outputs

Grid [raster]
<put output description here>

Console usage

processing.runalg('saga:modifedquadraticshepard', shapes, field, target, quadratic_neighbors, weighting_neighbors, user_xmin, user_xmax, user_ymin, user_ymax, user_size, user_grid)

See also

Natural neighbour

Description

<put algorithm description here>

Parameters

Points [vector: point]
<put parameter description here>
Attribute [tablefield: any]
<put parameter description here>
Target Grid [selection]

<put parameter description here>

Options:

  • 0 — [0] user defined

Default: 0

Sibson [boolean]

<put parameter description here>

Default: True

Output extent [extent]

<put parameter description here>

Default: 0,1,0,1

Cellsize [number]

<put parameter description here>

Default: 100.0

Outputs

Grid [raster]
<put output description here>

Console usage

processing.runalg('saga:naturalneighbour', shapes, field, target, sibson, output_extent, user_size, user_grid)

See also

Nearest neighbour

Description

<put algorithm description here>

Parameters

Points [vector: point]
<put parameter description here>
Attribute [tablefield: any]
<put parameter description here>
Target Grid [selection]

<put parameter description here>

Options:

  • 0 — [0] user defined

Default: 0

Output extent [extent]

<put parameter description here>

Default: 0,1,0,1

Cellsize [number]

<put parameter description here>

Default: 100.0

Outputs

Grid [raster]
<put output description here>

Console usage

processing.runalg('saga:nearestneighbour', shapes, field, target, output_extent, user_size, user_grid)

See also

Shapes to grid

Description

<put algorithm description here>

Parameters

Shapes [vector: any]
<put parameter description here>
Attribute [tablefield: any]
<put parameter description here>
Method for Multiple Values [selection]

<put parameter description here>

Options:

  • 0 — [0] first
  • 1 — [1] last
  • 2 — [2] minimum
  • 3 — [3] maximum
  • 4 — [4] mean

Default: 0

Method for Lines [selection]

<put parameter description here>

Options:

  • 0 — [0] thin
  • 1 — [1] thick

Default: 0

Preferred Target Grid Type [selection]

<put parameter description here>

Options:

  • 0 — [0] Integer (1 byte)
  • 1 — [1] Integer (2 byte)
  • 2 — [2] Integer (4 byte)
  • 3 — [3] Floating Point (4 byte)
  • 4 — [4] Floating Point (8 byte)

Default: 0

Output extent [extent]

<put parameter description here>

Default: 0,1,0,1

Cellsize [number]

<put parameter description here>

Default: 100.0

Outputs

Grid [raster]
<put output description here>

Console usage

processing.runalg('saga:shapestogrid', input, field, multiple, line_type, grid_type, output_extent, user_size, user_grid)

See also

Triangulation

Description

<put algorithm description here>

Parameters

Points [vector: point]
<put parameter description here>
Attribute [tablefield: any]
<put parameter description here>
Target Grid [selection]

<put parameter description here>

Options:

  • 0 — [0] user defined

Default: 0

Output extent [extent]

<put parameter description here>

Default: 0,1,0,1

Cellsize [number]

<put parameter description here>

Default: 100.0

Outputs

Grid [raster]
<put output description here>

Console usage

processing.runalg('saga:triangulation', shapes, field, target, output_extent, user_size, user_grid)

See also