Class pv.Scale.linear
Extends
pv.Scale.quantitative.
Represents a linear scale; a function that performs a linear
transformation. Most
commonly, a linear scale represents a 1-dimensional linear transformation
from a numeric domain of input data [d0,
d1] to a numeric range of pixels [r0,
r1]. The equation for such a scale is:
f(x) = (x - d0) / (d1 - d0) * (r1 - r0) + r0For example, a linear scale from the domain [0, 100] to range [0, 640]:
f(x) = (x - 0) / (100 - 0) * (640 - 0) + 0Thus, saying
f(x) = x / 100 * 640
f(x) = x * 6.4
.height(function(d) d * 6.4)is identical to
.height(pv.Scale.linear(0, 100).range(0, 640))Note that the scale is itself a function, and thus can be used as a property directly, assuming that the data associated with a mark is a number. While this is convenient for single-use scales, frequently it is desirable to define scales globally:
var y = pv.Scale.linear(0, 100).range(0, 640);The y scale can now be equivalently referenced within a property:
.height(function(d) y(d))Alternatively, if the data are not simple numbers, the appropriate value can be passed to the y scale (e.g., d.foo). The #by method similarly allows the data to be mapped to a numeric value before performing the linear transformation.
Defined in: LinearScale.js.
| Constructor Attributes | Constructor Name and Description |
|---|---|
|
Returns a linear scale for the specified domain.
|
- Methods borrowed from class pv.Scale.quantitative:
- by, domain, invert, nice, range, tickFormat, ticks
Class Detail
pv.Scale.linear()
Returns a linear scale for the specified domain. The arguments to this
constructor are optional, and equivalent to calling #domain.
The default domain and range are [0,1].
- Parameters:
- {number...} domain...
- optional domain values.