Update CHANGELOG

Author: mbostock

CHANGELOG.md

@@ -22,6 +22,7 @@ The resulting object is:
 ({
   type: "linear",
   domain: [-0.78, 1.35],
+  range: [0, 1],
   interpolate: d3.interpolateTurbo,
   clamp: false,
   label: "Anomaly"
@@ -34,14 +35,12 @@ Inspecting materialized scales is useful for debugging and for generating color
 Plot.plot({…, color: myplot.scale("color")})
 ```
 
-The new *scale*.**unknown** option provides more control over how unexpected or undefined values are displayed. For quantitative and temporal scales, the unknown value is used whenever the input value is undefined, null, or NaN. For ordinal or categorical scales, the unknown value is returned for any input value outside the domain. For band or point scales, the unknown option has no effect; it is effectively always equal to undefined. If the unknown option is set to undefined (the default), or null or NaN, then the affected input values will be considered undefined and filtered from the output.
+The new *scale*.**unknown** option provides more control over how unexpected, undefined, or invalid values are displayed. For quantitative and temporal scales, the unknown value is returned whenever the input value is undefined, null, or NaN. For ordinal or categorical scales, the unknown value is returned for any input value outside the domain. For band or point scales, the unknown option has no effect; it is always equal to undefined. If the unknown option is set to undefined (the default), or null or NaN, then affected input values will be considered undefined and filtered from the output. Hence, by setting *scale*.unknown to a defined value, you can ensure that these values are plotted.
 
-Plot now supports piecewise (a.k.a. “polylinear”) scales. If a single-argument *interpolate* function is specified (such as d3.interpolateWarm, or equivalently if you specify a color *scheme* such as *warm*), and the *range* is undefined, and the *domain* has more than two elements, the range [0, 1] will be partitioned into *domain*.length - 1 same-sized segments. The default scale range for *x* and *y* now supports piecewise scales, as does the default axis tick count.
+Plot now supports piecewise (a.k.a. “polylinear”) scales. If a single-argument *interpolate* function is specified (such as d3.interpolateWarm, or equivalently if you specify a color *scheme* such as *warm*), and the *range* is undefined, and the *domain* has more than two elements, the range [0, 1] will be partitioned into *domain*.length - 1 equal-size segments. The default scale range for *x* and *y* now supports piecewise scales, as does the default axis tick count. Additionally, ordinal color schemes now return the correct number of colors when the natural scheme size is not equal to the desired scheme size; and the *rainbow* and *sinebow* cyclical color schemes, when used with an ordinal color scale, no longer duplicate the first color as the last color.
 
 Plot now automatically detects “reversed” quantitative or temporal scales with descending domains or ranges. Reversed scales are detected by comparing the natural order of the domain and range. This improves the default axis label arrow orientation and the behavior of the *zero* option.
 
-Ordinal color schemes now return the correct number of colors when the natural scheme size is not equal to the desired scheme size. The *rainbow* and *sinebow* cyclical color schemes, when used with an ordinal color scale, no longer duplicate the first color as the last color.
-
 The *interval* option for rect and bar now works with the default identity accessor. Also, the *interval* option is now ignored if the necessary value channel (*e.g.*, *x*) is not defined, rather than throwing an error. For example, you no longer have to define *x* as the identity function here:
 
 ```js