Benchmark: Comparing different techniques to truncate float-point numbers in JavaScript

Comparing different techniques to truncate float-point numbers in JavaScript (version: 0)

Math.trunc(original) ~~original; // Double negation original & -1; // Bitwise AND with -1 original | 0; // Bitwise OR with 0 original ^ 0; // Bitwise XOR with 0 original >> 0; // Bitwise shifting by 0

Comparing performance of: Math.trunc() vs Double negation vs Bitwise AND with -1 vs Bitwise OR with 0 vs Bitwise XOR with 0 vs Bitwise shifting by 0

Created: 2 years ago by: Guest

Jump to the latest result

Script Preparation code:

var original = 3.14;

Tests:

Math.trunc()
var truncated1 = Math.trunc(original)
Double negation
const truncated1 = ~~original;
Bitwise AND with -1
const truncated2 = original & -1;
Bitwise OR with 0
const truncated3 = original | 0;
Bitwise XOR with 0
const truncated4 = original ^ 0;
Bitwise shifting by 0
const truncated5 = original >> 0;

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
Math.trunc()
Double negation
Bitwise AND with -1
Bitwise OR with 0
Bitwise XOR with 0
Bitwise shifting by 0

Fastest: N/A

Slowest: N/A

Latest run results:

Run details: (Test run date: 2 years ago)

User agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.5112.102 Safari/537.36 Edg/104.0.1293.63

Browser/OS: Chrome 104 on Windows

View result in a separate tab

Test name	Executions per second
Math.trunc()	438931.0 Ops/sec
Double negation	1121380.8 Ops/sec
Bitwise AND with -1	1287382.6 Ops/sec
Bitwise OR with 0	1347989.9 Ops/sec
Bitwise XOR with 0	1277862.6 Ops/sec
Bitwise shifting by 0	1315309.4 Ops/sec

Autogenerated LLM Summary (model llama3.2:3b, generated 2 months ago):

LLMs can make mistakes. Check important info.

Let's dive into the world of JavaScript microbenchmarks and explore what's being tested on MeasureThat.net.

**Benchmark Definition**

The benchmark definition describes six different techniques to truncate float-point numbers in JavaScript:

1. `Math.trunc(original)`
2. Double negation (`~~original`)
3. Bitwise AND with -1 (`original & -1`)
4. Bitwise OR with 0 (`original | 0`)
5. Bitwise XOR with 0 (`original ^ 0`)
6. Bitwise shifting by 0 (`original >> 0`)

Each technique is represented as a separate benchmark definition, which includes the JavaScript code to perform the truncation and the name of the test case.

**Pros and Cons of Each Approach**

1. **Math.trunc(original)**: This method uses the built-in `trunc()` function, which is likely implemented in native code. It's fast and efficient but may not be as readable or intuitive for developers.
2. **Double negation (`~~original`)**: This method uses a clever trick by applying two consecutive negations to truncate the number. While it's concise, it may lead to confusion among developers who are not familiar with this idiom.
3. **Bitwise AND with -1 (`original & -1`)**: This method uses bitwise operations to clear the fractional part of the number. It's a common technique in C-style programming but might be less familiar to JavaScript developers.
4. **Bitwise OR with 0 (`original | 0`)**: Similar to the previous approach, this method uses bitwise operations to clear the fractional part of the number. However, it may not work as expected for negative numbers due to sign bit masking.
5. **Bitwise XOR with 0 (`original ^ 0`)**: This method is essentially equivalent to the previous one but might be slightly more efficient due to its simplicity.
6. **Bitwise shifting by 0 (`original >> 0`)**: This method simply shifts the number right by zero, which has no effect on the value. While it's a valid operation, it's unlikely to provide any benefits over other methods.

**Library Usage**

None of the benchmark definitions explicitly uses a JavaScript library. However, if we look at the `Math.trunc()` implementation, it's likely that the browser's JavaScript engine is using a native implementation of this function, which might be optimized for performance.

**Special JS Features or Syntax**

There are no special JavaScript features or syntax used in these benchmark definitions, aside from the double negation trick (`~~original`), which is an idiosyncratic feature of JavaScript. However, it's worth noting that some browsers may optimize certain operations more than others due to their specific implementation details.

**Alternatives**

Other alternatives for truncating float-point numbers in JavaScript include:

* Using a library like `decimal.js` or `Big.js`, which provide decimal arithmetic and rounding functions.
* Implementing a custom algorithm using bitwise operations, as shown in the benchmark definitions.
* Using a different programming paradigm, such as C++ or Rust, which might offer better performance for numerical computations.

In conclusion, the MeasureThat.net benchmark provides an interesting comparison of different techniques for truncating float-point numbers in JavaScript. While each method has its pros and cons, `Math.trunc(original)` is likely the fastest and most efficient approach due to its native implementation.

Related benchmarks:

Comparing different techniques to truncate float-point numbers in JavaScript (version: 0)

Math.trunc(original) ~~original; // Double negation original & -1; // Bitwise AND with -1 original | 0; // Bitwise OR with 0 original ^ 0; // Bitwise XOR with 0 original >> 0; // Bitwise shifting by 0

Comparing performance of: Math.trunc() vs Double negation vs Bitwise AND with -1 vs Bitwise OR with 0 vs Bitwise XOR with 0 vs Bitwise shifting by 0

Created: 2 years ago by: Guest

Jump to the latest result

Math.trunc()

Double negation

Bitwise AND with -1

Bitwise OR with 0

Bitwise XOR with 0

Bitwise shifting by 0

Suite status: <idle, ready to run>

Experimental features:

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated 2 months ago):