quantlib

QuantLib Async Bindings for Node.js

This open source project brings QuantLib to the Node.js community, it's similar to QuantLibXL project which is for Microsoft Excel.

Most functions in QuantLibXL can be used in the similar way in Node on the server side.

All functions in this project are Async, they are exported to Promise sytle function, please see Example below.

Getting started

npm install quantlib

• Windows (32-bit & 64-bit)

npm install quantlib will do everything, including the node package installation and pre-built native addon (no dependency) download, you can start use it right away.

v0.2.x windows prebuilt on windows 8.1 with VC14

v0.1.x windows prebuilt on windows 7 with VC12, only 32-bit available

• Linux & Mac

Please refer to how to build below

QuantLib Documents

• Function Categories

• Function List

Version Matrix

QuantLib	QuantLibAddin	Node.js	quantlib.node
1.7.1	1.7.0	6.9.1	0.1.x
1.8.1	1.8.0	6.9.5	0.2.x
1.9.2	1.9.0	6.9.5	0.3.x

Building the native addon

Prerequisite

• CMake 2.8 or above, Visual C++ for windows, Xcode for Mac, GCC for Linux
• Node.js according to version matrix
• nan ^2.2.0
• node-gyp ~3.0.3
• QuantLib, QuantLibAddin, ObjectHandler source code according to version matrix, they need to be put in the same directory
• boost - which is required to build QuantLib

for Mac OS X and Linux, QuantLib projects must be built with -std=c++11 this flag is required by node

for linux, need to add -fPIC flag

Set environment variable

please refer to cmake/*.cmake and CMakeList.txt files

• NAN_DIR - location of nan
• NODE_GYP_DIR - location of .node-gyp generated by node-gyp tool, which is at ~/.node-gyp, if it doesn't exist, follow the instruction in node-gyp, and build a helloword program, it will generate the .node-gyp directory
• QUANTLIB_ROOT - location of QuantLib, QuantLibAddin, ObjectHandler source code
• BOOST_ROOT - location boost installed

Use cmake to build the addon

• Build QuantLib and QuantLibAddin, please check CMakeList.txt for library name, and make sure generated library names are the same in CMakeList.txt
• from quantlibnode root directory cd build
• cmake .. for Windows and Linux, cmake -G Xcode .. for Mac OS X
• cmake --build . --config Release
• For Linux, you may need to put quantlib.node under build/Release manually,

Example

This is from QuantLibXL - StandaloneExamples - Math - RankReduction.xlsx - FirstCorrMatrix

C2 - formulat: =qlSymmetricSchurDecomposition(,B3:K12)

M3:12 - forumula: {=qlSymmetricSchurDecompositionEigenvalues(C2)}

the input/output are the same

var ql = require('quantlib');

var mtx1 =
[
  [1.00000,	0.97560,	0.95240,	0.93040,	0.90940,	0.88940,	0.87040,	0.85230,	0.83520,	0.81880],
  [0.97560,	1.00000,    0.97560,    0.95240,    0.93040,    0.90940,    0.88940,    0.87040,    0.85230,    0.83520],
  [0.95240,	0.97560,	1.00000,	0.97560,	0.95240,	0.93040,	0.90940,	0.88940,	0.87040,	0.85230],
  [0.93040,	0.95240,	0.97560,	1.00000,	0.97560,	0.95240,	0.93040,	0.90940,	0.88940,	0.87040],
  [0.90940,	0.93040,	0.95240,	0.97560,	1.00000,	0.97560,	0.95240,	0.93040,	0.90940,	0.88940],
  [0.88940,	0.90940,	0.93040,	0.95240,	0.97560,	1.00000,	0.97560,	0.95240,	0.93040,	0.90940],
  [0.87040,	0.88940,	0.90940,	0.93040,	0.95240,	0.97560,	1.00000,	0.97560,	0.95240,	0.93040],
  [0.85230,	0.87040,	0.88940,	0.90940,	0.93040,	0.95240,	0.97560,	1.00000,	0.97560,	0.95240],
  [0.83520,	0.85230,	0.87040,	0.88940,	0.90940,	0.93040,	0.95240,	0.97560,	1.00000,	0.97560],
  [0.81880,	0.83520,	0.85230,	0.87040,	0.88940,	0.90940,	0.93040,	0.95240,	0.97560,	1.00000]
];

/* 
 * call Constructor to create object, objectID needs to be provided as input, this is diffierent with QuantLibXL
 * ObjectID, can be generated with Node.js, it needs to be uniquie
 * since this is on server side, you may put session# as part of objectID
*/
ql.SymmetricSchurDecomposition('mtx#1',mtx1).then(function(obj){ //C2 formula: =qlSymmetricSchurDecomposition(,B3:K12)
  
  // call Method with objectID, this is same as QuantLibXL
  ql.SymmetricSchurDecompositionEigenvalues(obj).then(function(r){ // M3:12 formula: {=qlSymmetricSchurDecompositionEigenvalues(C2)}
    console.log(r);
  });

}).catch(function(e){
  console.log(e);
});

>
[ 9.270906840163782,
  0.4207173234885105,
  0.12674770658244172,
  0.059239731356788505,
  0.03595303870722261,
  0.024956978505270924,
  0.019117669503864024,
  0.01580103250921176,
  0.01377474504269164,
  0.012784934140218302 ]