node-topography

canvas-topography =================

npm install node-topography
8 downloads in the last month

canvas-topography

A Node module for manipulating 2-d data; it is intended as a library for canvas data manipulation, but any scenario where 2d manipulation of content is required might find this useful.

Data is stored in a flat array, accessed by interpolation of x-y based on width and height of the dataset. Data can be any sort of format you like - numeric, objects, strings, etc.

The node-topography-canvas library requires data to be arrays of four numbers 0..255 (pref. integral). However, intermediary data states, such as integral data, are required to drive filters in certain scenarios.

API

factory(mixins, config)

The factory initializes and returns a new TopoGrid. The form and digestion pattern of the input depends on the values of config.

  • If you want to insert array data into the grid directly, assert {source: , source_type: 'array', width: , height: }

  • You can also assert a "data factory function" that takes (x, y) and returns a value. It will be bound to the TopoGrid.

  • You can load in images asnd canvas using the node-topography-canvas mixin as well.

neighbors(x, y, range) and neighbor_values(x, y, range, asInt)

Neighbors returns an array of data within +/- range; only "good" data (data for coordinates on the grid) are returned.

For each value an object is returned:

{
    x:  x_value,
    y: y_value,
    distance:      self.distance(x, y, x_value, y_value),
    orth_distance: self.orth_distance(x, y, x_value, y_value),
    x_offset:      x_value - x,
    y_offset:      y_value - y,
    value:         xy_value,
    rise:          xy_value - value // if both values are numeric
}

Other than rise, all the properties are related to the x, y coordinates and don't depend on the actual value being any specific format.

Neighbor values return a pure array of the values; while there is no explicit data or promise of which value came from what x,y coordinate, they are currently returned in x, y order. This is best for when positionless statistics on a group of data is required.

the "ring" parameter forces the neighbor method to return boundry values only - so a neighbor(3, 4, 2) would return data from point (2,3), but neighbor (3, 4, 2, true) would not.

Combine(cb, other_topo, combiner_filter, clone)

Combine merges two TopoGrids. If clone is true, the combination is returned as a third, new topo; if not, the combination overwrites the first ( calling) grid. The grids are assumed to be the same size.

The combine_filter is bound to the first grid and has the api(gridA_value, gridB_value, x, y); the result of the combinder_filter is asserted into the clone, or the first grid.

xy(x, y)

returns the data array index of a coordinate pair. throws an error for off-grid points.

deindex(index)

returns a two-value array for a given data array offset.

value(x, y) | value(xy_array)

returns the data at a given coordinate

xy_good(x, y)

returns a boolean for whether or not the point is in the 0...width, 0..height range.

compress_to_greyscale(method)

compresses image data to a single value in the 0..255 range. How it does so depends on input:

'r', 'g', 'b', 'a' will take a single channel and get that value.

'avg' (or no input) will average the first three array values.

note, in most cases, opacity is ignored.

npm loves you