exclusiveFacets (observablehq#1649)

* exclusiveFacets

* pad the data with duplicates

* reindex

* test

* done

* reindex iterables

* reindex symbol

---------

Co-authored-by: Philippe Rivière <[email protected]>

src/options.js

@@ -7,6 +7,10 @@ import {maybeTimeInterval, maybeUtcInterval} from "./time.js";
 export const TypedArray = Object.getPrototypeOf(Uint8Array);
 const objectToString = Object.prototype.toString;
 
+// If a reindex is attached to the data, channel values expressed as arrays will
+// be reindexed when the channels are instantiated. See exclusiveFacets.
+export const reindex = Symbol("reindex");
+
 export function valueof(data, value, type) {
   const valueType = typeof value;
   return valueType === "string"
@@ -17,7 +21,11 @@ export function valueof(data, value, type) {
     ? map(data, constant(value), type)
     : typeof value?.transform === "function"
     ? maybeTypedArrayify(value.transform(data), type)
-    : maybeTypedArrayify(value, type);
+    : maybeTake(maybeTypedArrayify(value, type), data?.[reindex]);
+}
+
+function maybeTake(values, index) {
+  return index ? take(values, index) : values;
 }
 
 function maybeTypedMap(data, f, type) {
@@ -170,6 +178,7 @@ export function isScaleOptions(option) {
 
 // Disambiguates an options object (e.g., {y: "x2"}) from a channel value
 // definition expressed as a channel transform (e.g., {transform: …}).
+// TODO Check typeof option[Symbol.iterator] !== "function"?
 export function isOptions(option) {
   return isObject(option) && typeof option.transform !== "function";
 }
@@ -223,7 +232,7 @@ export function where(data, test) {
 
 // Returns an array [values[index[0]], values[index[1]], …].
 export function take(values, index) {
-  return map(index, (i) => values[i]);
+  return map(index, (i) => values[i], values.constructor);
 }
 
 // If f does not take exactly one argument, wraps it in a function that uses take.

src/transforms/exclusiveFacets.js

@@ -0,0 +1,40 @@
+import {reindex, slice} from "../options.js";
+
+export function exclusiveFacets(data, facets) {
+  if (facets.length === 1) return {data, facets}; // only one facet; trivially exclusive
+
+  const n = data.length;
+  const O = new Uint8Array(n);
+  let overlaps = 0;
+
+  // Count the number of overlapping indexes across facets.
+  for (const facet of facets) {
+    for (const i of facet) {
+      if (O[i]) ++overlaps;
+      O[i] = 1;
+    }
+  }
+
+  // Do nothing if the facets are already exclusive.
+  if (overlaps === 0) return {data, facets}; // facets are exclusive
+
+  // For each overlapping index (duplicate), assign a new unique index at the
+  // end of the existing array, duplicating the datum. For example, [[0, 1, 2],
+  // [2, 1, 3]] would become [[0, 1, 2], [4, 5, 3]]. Also attach a reindex to
+  // the data to preserve the association of channel values specified as arrays.
+  data = slice(data);
+  const R = (data[reindex] = new Uint32Array(n + overlaps));
+  facets = facets.map((facet) => slice(facet, Uint32Array));
+  let j = n;
+  O.fill(0);
+  for (const facet of facets) {
+    for (let k = 0, m = facet.length; k < m; ++k) {
+      const i = facet[k];
+      if (O[i]) (facet[k] = j), (data[j] = data[i]), (R[j] = i), ++j;
+      else R[i] = i;
+      O[i] = 1;
+    }
+  }
+
+  return {data, facets};
+}

src/transforms/stack.js

@@ -4,6 +4,7 @@ import {withTip} from "../mark.js";
 import {maybeApplyInterval, maybeColumn, maybeZ, maybeZero} from "../options.js";
 import {column, field, mid, one, range, valueof} from "../options.js";
 import {basic} from "./basic.js";
+import {exclusiveFacets} from "./exclusiveFacets.js";
 
 export function stackX(stackOptions = {}, options = {}) {
   if (arguments.length === 1) [stackOptions, options] = mergeOptions(stackOptions);
@@ -85,6 +86,7 @@ function stack(x, y = one, kx, ky, {offset, order, reverse}, options) {
   order = maybeOrder(order, offset, ky);
   return [
     basic(options, (data, facets, plotOptions) => {
+      ({data, facets} = exclusiveFacets(data, facets));
       const X = x == null ? undefined : setX(maybeApplyInterval(valueof(data, x), plotOptions?.[kx]));
       const Y = valueof(data, y, Float64Array);
       const Z = valueof(data, z);