diff --git a/test/core/import.test.js b/test/core/import.test.js index dfc5eb408..f1d58416e 100644 --- a/test/core/import.test.js +++ b/test/core/import.test.js @@ -86,49 +86,6 @@ describe('import', function() { }); - it('should extend math with numbers', function() { - // extend math.js with numbers.js - // examples copied from https://github.com/sjkaliski/numbers.js/blob/master/examples/statistic.js - math.import(require('numbers'), {wrap: true, silent: true}); - - assert.equal(math.fibonacci(7), 13); - - // Consider a data representing total follower count of a - // variety of users. - var followers = math.matrix([100, 50, 1000, 39, 283, 634, 3, 6123]); - - // We can generate a report of summary statistics - // which includes the mean, 1st and 3rd quartiles, - // and standard deviation. - var report = math.report(followers); - approx.deepEqual(report, { - mean: 1029, - firstQuartile: 44.5, - median: 191.5, - thirdQuartile: 817, - standardDev: 1953.0897316815733 - }); - - // Maybe we decide to become a bit more curious about - // trends in follower count, so we start conjecturing about - // our ability to "predict" trends. - // Let's consider the number of tweets those users have. - var tweets = math.matrix([100, 10, 400, 5, 123, 24, 302, 2000]); - - // Let's calculate the correlation. - var correlation = math.correlation(tweets, followers); - approx.equal(correlation, 0.98054753183666); - - // Now let's create a linear regression. - var linReg = math.linearRegression(tweets, followers); - - // linReg is actually a function we can use to map tweets - // onto followers. We'll see that around 1422 followers - // are expected if a user tweets 500 times. - var estFollowers = linReg(500); - approx.equal(estFollowers, 1422.431464053916); - }); - it('should throw an error in case of wrong number of arguments', function () { assert.throws (function () {math.import()}, /ArgumentsError/); assert.throws (function () {math.import('', {}, 3)}, /ArgumentsError/);