refactor: extract computeScore method

GoogleChrome · Jul 18, 2017 · f0909fc · f0909fc
1 parent 76cdb54
commit f0909fc
Show file tree

Hide file tree

Showing 8 changed files with 59 additions and 62 deletions.
diff --git a/lighthouse-core/audits/audit.js b/lighthouse-core/audits/audit.js
@@ -5,6 +5,8 @@
  */
 'use strict';
 
+const statistics = require('../lib/statistics');
+
 const DEFAULT_PASS = 'defaultPass';
 
 class Audit {
@@ -32,6 +34,27 @@ class Audit {
     throw new Error('Audit meta information must be overridden.');
   }
 
+  /**
+   * Computes and clamps the score between 0 and 100 based on the measured value, and the two control
+   * points on the log-normal distribution (point of diminishing returns and the median value).
+   *
+   * @param {number} measuredValue
+   * @param {number} diminishingReturnsValue
+   * @param {number} medianValue
+   * @return {number}
+   */
+  static computeScore(measuredValue, diminishingReturnsValue, medianValue) {
+    const distribution = statistics.getLogNormalDistribution(
+      medianValue,
+      diminishingReturnsValue
+    );
+
+    let score = 100 * distribution.computeComplementaryPercentile(measuredValue);
+    score = Math.min(100, score);
+    score = Math.max(0, score);
+    return Math.round(score);
+  }
+
   /**
    * @param {!Audit} audit
    * @param {string} debugString

diff --git a/lighthouse-core/audits/byte-efficiency/total-byte-weight.js b/lighthouse-core/audits/byte-efficiency/total-byte-weight.js
@@ -9,7 +9,6 @@
 'use strict';
 
 const ByteEfficiencyAudit = require('./byte-efficiency-audit');
-const statistics = require('../../lib/statistics');
 
 // Parameters for log-normal CDF scoring. See https://www.desmos.com/calculator/gpmjeykbwr
 // ~75th and ~90th percentiles http://httparchive.org/interesting.php?a=All&l=Feb%201%202017&s=All#bytesTotal
@@ -72,9 +71,11 @@ class TotalByteWeight extends ByteEfficiencyAudit {
       //   <= 1600KB: score≈100
       //   4000KB: score=50
       //   >= 9000KB: score≈0
-      const distribution = statistics.getLogNormalDistribution(
-        SCORING_MEDIAN, SCORING_POINT_OF_DIMINISHING_RETURNS);
-      const score = 100 * distribution.computeComplementaryPercentile(totalBytes);
+      const score = ByteEfficiencyAudit.computeScore(
+        totalBytes,
+        SCORING_POINT_OF_DIMINISHING_RETURNS,
+        SCORING_MEDIAN
+      );
 
       const headings = [
         {key: 'url', itemType: 'url', text: 'URL'},
@@ -85,10 +86,10 @@ class TotalByteWeight extends ByteEfficiencyAudit {
       const tableDetails = ByteEfficiencyAudit.makeTableDetails(headings, results);
 
       return {
+        score,
         rawValue: totalBytes,
         optimalValue: this.meta.optimalValue,
         displayValue: `Total size was ${ByteEfficiencyAudit.bytesToKbString(totalBytes)}`,
-        score: Math.round(Math.max(0, Math.min(score, 100))),
         extendedInfo: {
           value: {
             results,

diff --git a/lighthouse-core/audits/consistently-interactive.js b/lighthouse-core/audits/consistently-interactive.js
@@ -8,7 +8,6 @@
 const Audit = require('./audit');
 const Util = require('../report/v2/renderer/util.js');
 const TracingProcessor = require('../lib/traces/tracing-processor');
-const statistics = require('../lib/statistics');
 
 // Parameters (in ms) for log-normal CDF scoring. To see the curve:
 //   https://www.desmos.com/calculator/uti67afozh
@@ -19,11 +18,6 @@ const REQUIRED_QUIET_WINDOW = 5000;
 const ALLOWED_CONCURRENT_REQUESTS = 2;
 const IGNORED_NETWORK_SCHEMES = ['data', 'ws'];
 
-const distribution = statistics.getLogNormalDistribution(
-  SCORING_MEDIAN,
-  SCORING_POINT_OF_DIMINISHING_RETURNS
-);
-
 /**
  * @fileoverview This audit identifies the time the page is "consistently interactive".
  * Looks for the first period of at least 5 seconds after FMP where both CPU and network were quiet,
@@ -234,14 +228,8 @@ class ConsistentlyInteractiveMetric extends Audit {
         const timeInMs = (timestamp - traceOfTab.timestamps.navigationStart) / 1000;
         const extendedInfo = Object.assign(quietPeriodInfo, {timestamp, timeInMs});
 
-        let score = 100 * distribution.computeComplementaryPercentile(timeInMs);
-        // Clamp the score to 0 <= x <= 100.
-        score = Math.min(100, score);
-        score = Math.max(0, score);
-        score = Math.round(score);
-
         return {
-          score,
+          score: Audit.computeScore(timeInMs, SCORING_POINT_OF_DIMINISHING_RETURNS, SCORING_MEDIAN),
           rawValue: timeInMs,
           displayValue: Util.formatMilliseconds(timeInMs),
           optimalValue: this.meta.optimalValue,

diff --git a/lighthouse-core/audits/dobetterweb/dom-size.js b/lighthouse-core/audits/dobetterweb/dom-size.js
@@ -13,7 +13,6 @@
 'use strict';
 
 const Audit = require('../audit');
-const statistics = require('../../lib/statistics');
 const Util = require('../../report/v2/renderer/util.js');
 
 const MAX_DOM_NODES = 1500;
@@ -73,12 +72,11 @@ class DOMSize extends Audit {
     //   <= 1500: score≈100
     //   3000: score=50
     //   >= 5970: score≈0
-    const distribution = statistics.getLogNormalDistribution(
-        SCORING_MEDIAN, SCORING_POINT_OF_DIMINISHING_RETURNS);
-    let score = 100 * distribution.computeComplementaryPercentile(stats.totalDOMNodes);
-
-    // Clamp the score to 0 <= x <= 100.
-    score = Math.max(0, Math.min(100, score));
+    const score = Audit.computeScore(
+      stats.totalDOMNodes,
+      SCORING_POINT_OF_DIMINISHING_RETURNS,
+      SCORING_MEDIAN
+    );
 
     const cards = [{
       title: 'Total DOM Nodes',
@@ -97,9 +95,9 @@ class DOMSize extends Audit {
     }];
 
     return {
+      score,
       rawValue: stats.totalDOMNodes,
       optimalValue: this.meta.optimalValue,
-      score: Math.round(score),
       displayValue: `${Util.formatNumber(stats.totalDOMNodes)} nodes`,
       extendedInfo: {
         value: cards

diff --git a/lighthouse-core/audits/estimated-input-latency.js b/lighthouse-core/audits/estimated-input-latency.js
@@ -8,7 +8,6 @@
 const Audit = require('./audit');
 const Util = require('../report/v2/renderer/util.js');
 const TracingProcessor = require('../lib/traces/tracing-processor');
-const statistics = require('../lib/statistics');
 
 // Parameters (in ms) for log-normal CDF scoring. To see the curve:
 // https://www.desmos.com/calculator/srv0hqhf7d
@@ -51,12 +50,14 @@ class EstimatedInputLatency extends Audit {
     //  Median = 100ms
     //  75th Percentile ≈ 133ms
     //  95th Percentile ≈ 199ms
-    const distribution = statistics.getLogNormalDistribution(SCORING_MEDIAN,
-        SCORING_POINT_OF_DIMINISHING_RETURNS);
-    const score = 100 * distribution.computeComplementaryPercentile(ninetieth.time);
+    const score = Audit.computeScore(
+      ninetieth.time,
+      SCORING_POINT_OF_DIMINISHING_RETURNS,
+      SCORING_MEDIAN
+    );
 
     return {
-      score: Math.round(score),
+      score,
       optimalValue: EstimatedInputLatency.meta.optimalValue,
       rawValue,
       displayValue: Util.formatMilliseconds(rawValue, 1),

diff --git a/lighthouse-core/audits/first-interactive.js b/lighthouse-core/audits/first-interactive.js
@@ -7,18 +7,12 @@
 
 const Audit = require('./audit');
 const Util = require('../report/v2/renderer/util.js');
-const statistics = require('../lib/statistics');
 
 // Parameters (in ms) for log-normal CDF scoring. To see the curve:
 //   https://www.desmos.com/calculator/rjp0lbit8y
 const SCORING_POINT_OF_DIMINISHING_RETURNS = 1700;
 const SCORING_MEDIAN = 10000;
 
-const distribution = statistics.getLogNormalDistribution(
-  SCORING_MEDIAN,
-  SCORING_POINT_OF_DIMINISHING_RETURNS
-);
-
 class FirstInteractiveMetric extends Audit {
   /**
    * @return {!AuditMeta}
@@ -46,14 +40,12 @@ class FirstInteractiveMetric extends Audit {
     const trace = artifacts.traces[Audit.DEFAULT_PASS];
     return artifacts.requestFirstInteractive(trace)
       .then(firstInteractive => {
-        let score = 100 * distribution.computeComplementaryPercentile(firstInteractive.timeInMs);
-        // Clamp the score to 0 <= x <= 100.
-        score = Math.min(100, score);
-        score = Math.max(0, score);
-        score = Math.round(score);
-
         return {
-          score,
+          score: Audit.computeScore(
+            firstInteractive.timeInMs,
+            SCORING_POINT_OF_DIMINISHING_RETURNS,
+            SCORING_MEDIAN
+          ),
           rawValue: firstInteractive.timeInMs,
           displayValue: Util.formatMilliseconds(firstInteractive.timeInMs),
           extendedInfo: {

diff --git a/lighthouse-core/audits/first-meaningful-paint.js b/lighthouse-core/audits/first-meaningful-paint.js
@@ -7,7 +7,6 @@
 
 const Audit = require('./audit');
 const Util = require('../report/v2/renderer/util.js');
-const statistics = require('../lib/statistics');
 
 // Parameters (in ms) for log-normal CDF scoring. To see the curve:
 // https://www.desmos.com/calculator/joz3pqttdq
@@ -101,17 +100,15 @@ class FirstMeaningfulPaint extends Audit {
     //   4000ms: score=50
     //   >= 14000ms: score≈0
     const firstMeaningfulPaint = traceOfTab.timings.firstMeaningfulPaint;
-    const distribution = statistics.getLogNormalDistribution(SCORING_MEDIAN,
-        SCORING_POINT_OF_DIMINISHING_RETURNS);
-    let score = 100 * distribution.computeComplementaryPercentile(firstMeaningfulPaint);
-
-    // Clamp the score to 0 <= x <= 100.
-    score = Math.min(100, score);
-    score = Math.max(0, score);
+    const score = Audit.computeScore(
+      firstMeaningfulPaint,
+      SCORING_POINT_OF_DIMINISHING_RETURNS,
+      SCORING_MEDIAN
+    );
 
     return {
+      score,
       duration: firstMeaningfulPaint.toFixed(1),
-      score: Math.round(score),
       rawValue: firstMeaningfulPaint,
       extendedInfo
     };

diff --git a/lighthouse-core/audits/speed-index-metric.js b/lighthouse-core/audits/speed-index-metric.js
@@ -6,7 +6,6 @@
 'use strict';
 
 const Audit = require('./audit');
-const statistics = require('../lib/statistics');
 const Util = require('../report/v2/renderer/util');
 
 // Parameters (in ms) for log-normal CDF scoring. To see the curve:
@@ -63,13 +62,11 @@ class SpeedIndexMetric extends Audit {
       //  Median = 5,500
       //  75th Percentile = 8,820
       //  95th Percentile = 17,400
-      const distribution = statistics.getLogNormalDistribution(SCORING_MEDIAN,
-        SCORING_POINT_OF_DIMINISHING_RETURNS);
-      let score = 100 * distribution.computeComplementaryPercentile(speedline.perceptualSpeedIndex);
-
-      // Clamp the score to 0 <= x <= 100.
-      score = Math.min(100, score);
-      score = Math.max(0, score);
+      const score = Audit.computeScore(
+        speedline.perceptualSpeedIndex,
+        SCORING_POINT_OF_DIMINISHING_RETURNS,
+        SCORING_MEDIAN
+      );
 
       const extendedInfo = {
         timings: {
@@ -97,7 +94,7 @@ class SpeedIndexMetric extends Audit {
       const rawValue = Math.round(speedline.perceptualSpeedIndex);
 
       return {
-        score: Math.round(score),
+        score,
         rawValue,
         displayValue: Util.formatNumber(rawValue),
         optimalValue: this.meta.optimalValue,