mueller
Simple demonastration of Muller's Recurrence problem which shows the floating point roundoff errors.
Use
You can get this example via NPM
npm install mueller
This example is implemented using in 3 ways:
- using native Number class
- using bignumber.js
- using decimal.js
In order to run it this example use the following command
node index.js [library] [decimal_places]
library can be "dec" or "big" for decimal.js or bignumber.js respectively. If it isn't provided native Number is used.
decimal_places is used for bignumber.js and represents the maximum number of decimal places of the results of operations. (default values is 20)
Benchmarks
The tests were done using Intel® Core™ i7-4702MQ @ 2.20 Ghz with default number of decimal places for bignumber.js
native
real 0m0.299suser 0m0.000ssys 0m0.031s
bignumber.js
real 0m28.282suser 0m0.000ssys 0m0.015s
decimal.js
real 0m25.758suser 0m0.015ssys 0m0.015s