222 lines
5.8 KiB
JavaScript
222 lines
5.8 KiB
JavaScript
'use strict';
|
|
|
|
function innerProduct(values1, values2) {
|
|
var result = 0;
|
|
for (var feature in values1) {
|
|
result += values1[feature] * (values2[feature] || 0.0);
|
|
}
|
|
|
|
return result;
|
|
}
|
|
|
|
function searchData(queryParams, minScore) {
|
|
var results = [];
|
|
|
|
for (var i = 0, count = DATA_RECORDS.length; i < count; ++i) {
|
|
var record = DATA_RECORDS[i];
|
|
var score = innerProduct(queryParams, record['rating']);
|
|
|
|
if (score >= minScore) {
|
|
results.push({
|
|
name: record['name'],
|
|
url: 'http://www.tripadvisor.com' + record['relativeUrl'],
|
|
score: score
|
|
});
|
|
}
|
|
}
|
|
|
|
results.sort(function(a, b) {
|
|
return b.score - a.score;
|
|
});
|
|
|
|
return results;
|
|
}
|
|
|
|
function searchStepper(range, steps, callback) {
|
|
var stepSize = range.getLength() / steps;
|
|
|
|
for (var i = 0; i < steps; ++i) {
|
|
var stepMax = range.end - stepSize * i;
|
|
var stepMin = stepMax - stepSize;
|
|
var stepMid = (stepMin + stepMax) / 2;
|
|
|
|
callback(stepMid);
|
|
}
|
|
}
|
|
|
|
function searchProjection(queryParams, minScore, feature, range, steps) {
|
|
var testParams = _.clone(queryParams);
|
|
var results = [];
|
|
|
|
searchStepper(range, steps, function(position) {
|
|
testParams[feature] = position;
|
|
results.push({
|
|
'sample': position,
|
|
'values': searchData(testParams, minScore)
|
|
});
|
|
});
|
|
|
|
return results;
|
|
}
|
|
|
|
function searchProjection2d(queryParams, minScore, feature1, feature2, range, steps) {
|
|
var testParams = _.clone(queryParams);
|
|
var results = [];
|
|
|
|
searchStepper(range, steps, function(sampleX) {
|
|
testParams[feature1] = sampleX;
|
|
searchStepper(range, steps, function(sampleY) {
|
|
testParams[feature2] = sampleY;
|
|
results.push({
|
|
'sampleX': sampleX,
|
|
'sampleY': sampleY,
|
|
'values': searchData(testParams, minScore)
|
|
});
|
|
});
|
|
});
|
|
|
|
return results;
|
|
}
|
|
|
|
function searchBuildHints(queryParams, minScore, feature, range, steps) {
|
|
var projection = searchProjection(
|
|
queryParams,
|
|
minScore,
|
|
feature,
|
|
range,
|
|
steps
|
|
);
|
|
|
|
var hints = [];
|
|
_.each(projection, function(result) {
|
|
hints.push({
|
|
'sample': result.sample,
|
|
'count': result.values.length
|
|
});
|
|
});
|
|
|
|
return hints;
|
|
}
|
|
|
|
function searchBuildHints2d(queryParams, minScore, feature1, feature2, range, steps) {
|
|
var projection = searchProjection2d(
|
|
queryParams,
|
|
minScore,
|
|
feature1,
|
|
feature2,
|
|
range,
|
|
steps
|
|
);
|
|
|
|
var hints = [];
|
|
_.each(projection, function(result) {
|
|
hints.push({
|
|
'sampleX': result.sampleX,
|
|
'sampleY': result.sampleY,
|
|
'count': result.values.length
|
|
});
|
|
});
|
|
|
|
return hints;
|
|
}
|
|
|
|
function outputResults(results, maxResults) {
|
|
$('#results').empty();
|
|
$('#count').text(results.length);
|
|
|
|
results = results.splice(0, maxResults);
|
|
|
|
var template = Handlebars.compile($('#template').html());
|
|
$('#results').append(template({'results': results}));
|
|
}
|
|
|
|
function onAdjust(name, value) {
|
|
var wa = window.adjuster;
|
|
var wg = window.grapher;
|
|
|
|
wa.queryParams[name] = value;
|
|
console.log(wa.queryParams);
|
|
|
|
var hintData = {};
|
|
_.each(wg.getColumnNames(), function(name) {
|
|
hintData[name] = searchBuildHints(wa.queryParams, wa.minScore, name, wa.searchRange, wa.hintSteps);
|
|
});
|
|
wg.setColumnHints(hintData);
|
|
|
|
var results = searchData(wa.queryParams, wa.minScore);
|
|
outputResults(results, wa.maxResults);
|
|
}
|
|
|
|
function onQuery() {
|
|
var query = $('#query').val();
|
|
var minScore = parseInt($('#minScore').val()) || 1.0;
|
|
var hintSteps = parseInt($('#hintSteps').val()) || 20;
|
|
var maxResults = parseInt($('#maxResults').val()) || 100;
|
|
var useLocalScale = true;
|
|
var useRelativeScale = true;
|
|
|
|
console.assert(query in DATA_KEYWORDS);
|
|
|
|
var queryParams = DATA_KEYWORDS[query];
|
|
var searchRange = new goog.math.Range(-1.0, 1.0);
|
|
var graphColumns = {};
|
|
|
|
for (var feature in queryParams) {
|
|
var hints = searchBuildHints(
|
|
queryParams,
|
|
minScore,
|
|
feature,
|
|
searchRange,
|
|
hintSteps
|
|
);
|
|
|
|
graphColumns[feature] = {
|
|
'color': '#607080',
|
|
'value': queryParams[feature],
|
|
'hints': hints,
|
|
'steps': hintSteps
|
|
}
|
|
}
|
|
|
|
window.adjuster = {
|
|
queryParams: queryParams,
|
|
searchRange: searchRange,
|
|
hintSteps: hintSteps,
|
|
minScore: minScore,
|
|
maxResults: maxResults
|
|
};
|
|
|
|
window.grapher = new Grapher('grapher', searchRange, useLocalScale, useRelativeScale);
|
|
window.grapher.setColumns(graphColumns);
|
|
window.grapher.setValueChangedListener(onAdjust);
|
|
|
|
var results = searchData(queryParams, minScore);
|
|
outputResults(results, maxResults);
|
|
|
|
$('#keyword').text(query);
|
|
$('#useLocalScale').prop('checked', useLocalScale);
|
|
$('#useRelativeScale').prop('checked', useRelativeScale);
|
|
$('#useLocalScale').click(function() {
|
|
var useLocalScale = $('#useLocalScale').is(':checked');
|
|
window.grapher.setUseLocalScale(useLocalScale);
|
|
});
|
|
$('#useRelativeScale').click(function() {
|
|
var useRelativeScale = $('#useRelativeScale').is(':checked');
|
|
window.grapher.setUseRelativeScale(useRelativeScale);
|
|
});
|
|
$('#input').fadeOut(function() {
|
|
$('#output').fadeIn();
|
|
});
|
|
}
|
|
|
|
$(document).ready(function() {
|
|
for (var keyword in DATA_KEYWORDS) {
|
|
$('#query').append($('<option></option>', {
|
|
'value': keyword,
|
|
'text': keyword
|
|
}));
|
|
}
|
|
|
|
$('#search').click(onQuery);
|
|
});
|