perform nearest-neighbor-search to match hexvals -> colornames

bin/preprocess-macrostrat-dataset.js

+#!/usr/bin/env node
+const data = require("../data/macrotstrat-lithclasses-all.json")
+const tinycolor = require("tinycolor2")
+const KNN = require("ml-knn")
+
+let train_dataset = Object.entries(tinycolor.names)
+  .map((c, k) => ({
+    color: c[0],
+    hex: c[1],
+    rgb: Object.values(tinycolor(c[1]).toRgb()).slice(0, 3),
+  }))
+  .map((c) => [c.rgb].flat())
+
+let train_labels = Object.entries(tinycolor.names)
+  .map((c, k) => ({
+    color: c[0],
+    hex: c[1],
+    rgb: Object.values(tinycolor(c[1]).toRgb()).slice(0, 3),
+  }))
+  .map((c) => `#${c.hex} ${c.color}`)
+
+knn = new KNN(train_dataset, train_labels, { k: 2 })
+
+let test_dataset = data.success.data.map((d) => {
+  let v = Object.values(tinycolor(d.color).toRgb()).slice(0, 3)
+  return v
+})
+
+let closest_colors = knn.predict(test_dataset)
+// console.log(
+//   closest_colors.map((v, k) => ({
+//     original: data[k].color,
+//     predicted: v,
+//   }))
+// )
+
+data.success.data.forEach((v, k) => {
+  v.color_predicted = closest_colors[k]
+  v.color_hex = closest_colors[k].split(" ")[0]
+  v.color_name = closest_colors[k].split(" ")[1]
+})
+console.log(JSON.stringify(data))
+//console.log(closest_colors)

package-lock.json

@@ -43,6 +43,19 @@
       "resolved": "https://registry.npmjs.org/lodash/-/lodash-4.17.19.tgz",
       "integrity": "sha512-JNvd8XER9GQX0v2qJgsaN/mzFCNA5BRe/j8JN9d+tWyGLSodKQHKFicdwNYzWwI3wjRnaKPsGj1XkBjx/F96DQ=="
     },
+    "ml-distance-euclidean": {
+      "version": "2.0.0",
+      "resolved": "https://registry.npmjs.org/ml-distance-euclidean/-/ml-distance-euclidean-2.0.0.tgz",
+      "integrity": "sha512-yC9/2o8QF0A3m/0IXqCTXCzz2pNEzvmcE/9HFKOZGnTjatvBbsn4lWYJkxENkA4Ug2fnYl7PXQxnPi21sgMy/Q=="
+    },
+    "ml-knn": {
+      "version": "3.0.0",
+      "resolved": "https://registry.npmjs.org/ml-knn/-/ml-knn-3.0.0.tgz",
+      "integrity": "sha512-w7Jig4vW09OYMurlIRRJj8yL2O/835QJpxup+yW7QVYXSjixmanPtPQL+weiTYbaB/wWX8JilB+Wllk4mxvP5w==",
+      "requires": {
+        "ml-distance-euclidean": "^2.0.0"
+      }
+    },
     "moment": {
       "version": "2.27.0",
       "resolved": "https://registry.npmjs.org/moment/-/moment-2.27.0.tgz",
@@ -52,6 +65,11 @@
       "version": "2.0.0",
       "resolved": "https://registry.npmjs.org/ms/-/ms-2.0.0.tgz",
       "integrity": "sha1-VgiurfwAvmwpAd9fmGF4jeDVl8g="
+    },
+    "tinycolor2": {
+      "version": "1.4.1",
+      "resolved": "https://registry.npmjs.org/tinycolor2/-/tinycolor2-1.4.1.tgz",
+      "integrity": "sha1-9PrTM0R7wLB9TcjpIJ2POaisd+g="
     }
   }
 }

package.json

@@ -13,7 +13,9 @@
     "faker": "^4.1",
     "getopts": "^2.2.5",
     "lodash": "^4.17.19",
-    "moment": "^2.27"
+    "ml-knn": "^3.0.0",
+    "moment": "^2.27",
+    "tinycolor2": "^1.4.1"
   },
   "devDependencies": {}
 }

src/geology-inserter.js

@@ -13,8 +13,11 @@ const axios = require("axios")
 const axutil = require("./axios-util.js")
 const querystring = require("querystring")
 const getopts = require("getopts")
+const tinycolor = require("tinycolor2")
 const Util = require("./classes/fake-util.js")
-const macrostrat_data = require("../data/macrotstrat-lithclasses-all.json")
+//["lith_id","name","type","group","class","color","fill","t_units"]
+const litho_data = require("../data/macrotstrat-lithclasses-extended.json")
+  .success.data
 const LithologyGeology = require("./classes/fake-lithology.js")
 const util = new Util()
 
@@ -27,8 +30,7 @@ const options = getopts(process.argv.slice(2), {
     expedition: 10, //SAMP
   },
 })
-//["lith_id","name","type","group","class","color","fill","t_units"]
-const litho_data = macrostrat_data.success.data
+
 //let core_url = "/api/v1/form?name=core&per-page=1&sort=-id"
 //const core_info = options.infile ? JSON.parse(options.infile) : {}
 const ax = axios.create(config.endpoint)
@@ -234,12 +236,21 @@ Promise.all(queries_map.values())
                     let sec_len = sp.section_length
                       ? sp.section_length
                       : util.round(util.frac_above(0.5), 2)
+                    let color = lithounits_per_split[s]?.color || "red"
+                    color = tinycolor(color).toName()
+                      ? `${color} (${tinycolor(color).toName()})`
+                      : color
+                    let color_name =
+                      lithounits_per_split[s]?.color_name || "red"
+                    let color_hex =
+                      lithounits_per_split[s]?.color_hex || "#FF0000"
+                    let rock_name = lithounits_per_split[s]?.name || "weird"
                     let fake_lithology = new LithologyGeology(
                       Object.assign({}, geology_lithology_columns, {
                         section_split_id: sp.id,
                         top_depth: sp.top_depth,
                         bottom_depth: sp.bottom_depth,
-                        color: lithounits_per_split[s]?.color || "red",
+                        color: color_name,
                         litho_unit: s,
                         rock_class:
                           lithounits_per_split[s]?.class || "artificial",
@@ -250,7 +261,8 @@ Promise.all(queries_map.values())
                           util.frac_below(0.1) * sec_len * 100 + 1,
                           0
                         ),
-                        description: `Spl${sp.id
+                        composition: rock_name,
+                        description: `${color} (${color_hex}) Spl${sp.id
                           .toString()
                           .padStart(4, "0")}, U${s
                           .toString()
@@ -259,7 +271,7 @@ Promise.all(queries_map.values())
                         }, secLen ${util.round(sp.section_length, 1)}m`,
                       }),
                       sp, // current
-                      sp // previous
+                      sp // previous - not used
                     ).fake()
                     fake_lithology_data.push(Object.assign({}, fake_lithology))
                     i++