823 Commits

Author SHA1 Message Date
jos
20f3af05b3 Fixed markdown urls, updated an example 2017-01-23 21:56:14 +01:00
jos
ddb8fc4374 Released v3.9.0 2017-01-23 20:47:52 +01:00
jos
12c2444ead Merge branch 'differentiation_consistent_api' into develop
# Conflicts:
#	HISTORY.md
2017-01-22 20:03:09 +01:00
Adam Kelly
4608afecc6 Change from multiply to multiplyScalar with kron 2017-01-22 18:46:11 +00:00
Jos de Jong
c879acf9d3 Merge pull request #780 from ericman314/differentiation_consistent_api
Fixed derivative of exp(x)
2017-01-22 15:44:57 +01:00
jos
d722473b53 Fixed simplify not handling non-existing functions correctly 2017-01-22 15:44:14 +01:00
ericman314
aab56c8094 Fixed derivative of exp(x) 2017-01-22 07:35:22 -07:00
jos
f4b7855a84 Docs, examples, some fixes 2017-01-22 15:30:18 +01:00
jos
db67f4cb84 Removed method Node.simplify() (use function simplify instead) 2017-01-22 13:42:19 +01:00
jos
08ed2b4160 Updated docs 2017-01-21 20:45:14 +01:00
jos
91dc47c763 Made ES5 complient again 2017-01-21 20:38:36 +01:00
jos
5c5b83988e Updaed docs of Node.simplify 2017-01-21 20:35:20 +01:00
jos
0ae4dce80b Updated docs of simplify and derivative 2017-01-21 20:29:10 +01:00
jos
6a61dd6d3c Implemented option {simplify: boolean} for function derivative 2017-01-21 20:10:10 +01:00
Adam Kelly
b564c20f19 Add Support for Complex Numbers in kron 2017-01-20 22:51:01 +00:00
jos
71ae3ea076 Consistent API for both derivative and simplify (dropped transform) 2017-01-20 21:43:40 +01:00
Adam Kelly
44515ba7ed Add support for 1D Vectors in math.kron 2017-01-20 18:09:42 +00:00
jos
1583b455ca Merge branch 'develop' into algebraic_differentiation 2017-01-20 11:18:05 +01:00
Adam Kelly
6e63a112a2 Add Kronecker Product (Direct / Tensor Product)
Adds Kronecker Product as math.kron(x, y);
2017-01-17 22:31:42 +00:00
jos
29da2b0418 Released v3.8.1 2016-12-15 21:07:26 +01:00
Sebastian Ruhleder
fe52d51c0a Add median absolute deviation 2016-12-08 18:29:11 +01:00
jos
a168041374 Functions add and multiply now accept more than two arguments. See #739. 2016-11-18 15:06:37 +01:00
jos
f491873452 Merge branch 'develop' into algebraic_differentiation 2016-11-07 20:57:27 +01:00
tetslee
5a75a8c222 Simplify terms and factors seperated by inverse operators.
The simplify algorithm now applies rules one at a time.
2016-11-07 11:20:28 +10:30
jos
1ecfdf1a0e Implemented BigNumber support in function arg() 2016-11-05 19:52:22 +01:00
tetslee
23bf1031eb Improved constant folding in simplify. 2016-11-04 18:02:42 +10:30
tetslee
67b69b49c5 Improvements to simplify(). Also changes to the way simplify and derivative are used. 2016-11-02 16:26:22 +10:30
tetslee
12cdee2956 Merge branch 'algebraic_differentiation' into derivative 2016-10-25 13:06:37 +10:30
jos
4b458d1779 Removed redundant imports 2016-10-21 17:58:53 +02:00
jos
22508ed8eb Moved constants in erf to the bottom (doc generator picks the first comment from a function) 2016-10-21 17:57:39 +02:00
Pat Grasso
d2ae67cca3 erf(Infinity) should return 1, not NaN
Likewise, erf(-Infinity) should be -1. This is true for any really large
|x|...say, |x| > 2^53.
2016-10-20 15:54:02 -04:00
Pat Grasso
a71cba9a7d Add TODO comments for future considerations
Specifically, implementation of erfi() (complex error function) and
precision when x is a bignumber.
2016-10-20 15:33:10 -04:00
Pat Grasso
57f082f64d Use internal sign() (Math.sign only in ES2015+) 2016-10-20 15:33:10 -04:00
Pat Grasso
ee41e35b5c Fix erf(bignumber) -> bignumber to return proper value
Note: The precision for the bignumber returned may not be accurate, as
the `erf()` approximation only guarantees precision to a certain degree.
2016-10-20 15:33:10 -04:00
Pat Grasso
80fd73c115 Update submodule index files to include erf 2016-10-20 15:33:09 -04:00
Pat Grasso
6ae793edea Implement erf(x) using W. J. Cody's approximation
Translate the Fortran77 implementation of W. J. Cody's "Rational
Chebyshev Approximations for the Error Function".
Place in new function subdirectory, `function/special`, because this
function doesn't seem to fit into any other category.
2016-10-20 15:33:09 -04:00
Jos de Jong
cd12316ec8 Merge pull request #679 from woylie/master
add pickMultipleRandom function
2016-10-19 21:39:05 +02:00
Mathias Polligkeit
cf857a3a09 some more minor code format changes 2016-10-18 02:02:09 +02:00
Mathias Polligkeit
b5a9695746 code formatting 2016-10-18 01:45:06 +02:00
Mathias Polligkeit
78f0c942bd clean up distribution.js 2016-10-18 01:43:49 +02:00
Mathias Polligkeit
f81abbab25 add type validation 2016-10-18 01:43:19 +02:00
Mathias Polligkeit
239cabec37 Revert "add an intermediary typed function that will fail a test"
This reverts commit fc2b730e044d3ec5a93c010d789eac47768fccc2.
2016-10-18 01:27:18 +02:00
Mathias Polligkeit
94d102e758 Revert "add randomHelper function for duplicate logic of random/randomInt; fails"
This reverts commit c464010234c73144df142cdb6fb7d32d0533bdf3.
2016-10-18 01:27:13 +02:00
Mathias Polligkeit
bfcfeafd2e Revert "refactor distribution.js"
This reverts commit ed0df31293f92171c44d06d5bdc04d408b967402.
2016-10-18 01:25:08 +02:00
Pat Grasso
30ae9c071f Change dimensionality of vector returned by cross()
If one of the input vectors/matrices has a dimension greater than 1, the
output vector will be a 1x3 matrix. Otherwise, the result will be a flat
3-component vector. Closes #716
2016-10-12 01:20:59 -04:00
patgrasso
a7f5bb27ef Extend cross() to support n-d vectors. Closes #716
Squeeze arrays before attempting to cross them so that n-dimensional
vectors can be cross multiplied (as long as only one direction has a
series of 3 elements)
2016-09-25 21:22:40 -04:00
jos
924694eeda Improved docs on transpose function (see #706) 2016-09-20 21:07:23 +02:00
jos
c2d39cc0f6 Remove some commented console logs (see #709) 2016-09-12 20:11:48 +02:00
Thomas Brierley
0d1b3f0f6d Fix print options argument type
Add missing 'Object' type definition to third paramater. Fixes #700
2016-08-12 16:40:00 +01:00
jos
9c6069b3cf Fixed broken bundle files (dist/math.js, dist/math.min.js) 2016-08-08 20:16:16 +02:00