FastDoubleParser
This is a Java port of Daniel Lemire's fast_float.
https://github.com/fastfloat/fast_float
Usage:
import ch.randelshofer.fastdoubleparser.FastDoubleParser;
import ch.randelshofer.fastdoubleparser.FastFloatParser;
double d = FastDoubleParser.parseDouble("1.2345");
float f = FastFloatParser.parseFloat("1.2345");
Method parseDouble() takes a CharacterSequence. a char-array or a byte-array as argument. This way. you can
parse from a StringBuffer or an array without having to convert your input to a String. Parsing from an array is
faster. because the parser can process multiple characters at once using SIMD instructions.
When you clone the code repository from github. you can choose from the following branches:
mainThe code in this branch requires Java 17.java8The code in this branch requires Java 8.devThis code may or may not work. This code uses the experimental Vector API, and the Foreign Memory Access API, that are included in Java 19.
How to run the performance tests on a Mac:
- Install Java JDK 8 or higher. for example OpenJDK Java 18
- Install the XCode command line tools from Apple.
- Open the Terminal and execute the following commands:
Command sequence for Java 18 or higher:
git clone https://github.com/wrandelshofer/FastDoubleParser.git
cd FastDoubleParser
javac --enable-preview -source 18 -d out -encoding utf8 --module-source-path src/main/java --module ch.randelshofer.fastdoubleparser
javac --enable-preview -source 18 -d out -encoding utf8 -p out --module-source-path FastDoubleParserDemo/src/main/java --module ch.randelshofer.fastdoubleparserdemo
java --enable-preview -p out -m ch.randelshofer.fastdoubleparserdemo/ch.randelshofer.fastdoubleparserdemo.Main
java --enable-preview -p out -m ch.randelshofer.fastdoubleparserdemo/ch.randelshofer.fastdoubleparserdemo.Main data/canada.txt
Command sequence for Java 8 or higher:
git clone https://github.com/wrandelshofer/FastDoubleParser.git
cd FastDoubleParser
git checkout java8
mkdir out
javac -d out -encoding utf8 -sourcepath src/main/java/ch.randelshofer.fastdoubleparser src/main/java/ch.randelshofer.fastdoubleparser/**/*.java
javac -d out -encoding utf8 -cp out -sourcepath FastDoubleParserDemo/src/main/java/ch.randelshofer.fastdoubleparserdemo FastDoubleParserDemo/src/main/java/ch.randelshofer.fastdoubleparserdemo/**/*.java
java -cp out ch.randelshofer.fastdoubleparserdemo.Main
java -cp out ch.randelshofer.fastdoubleparserdemo.Main data/canada.txt
On my Mac mini (2018) I get the results shown below. The speedup factor with respect to Double.parseDouble ranges from
0.5 to 6 depending on the shape of the input data. You can expect a speedup factor of 4 for common input data shapes.
Intel(R) Core(TM) i7-8700B CPU @ 3.20GHz
x86_64, Mac OS X, 12.5.1, 12
OpenJDK 64-Bit Server VM, Oracle Corporation, 17+35-2724
-XX:+UnlockDiagnosticVMOptions, -XX:CompileCommand=inline,java/lang/String.charAt
Parsing random doubles in the range [0,1).
[...]
Measuring: Trying to reach a confidence level of 99.8 % which only deviates by 1 % from the average measured duration.
[...]
Measurement results:
java.lang.Double : 87.56 MB/s (+/- 6.7 %) 5.03 Mfloat/s 198.97 ns/f
java.lang.Float : 92.64 MB/s (+/- 5.7 %) 5.32 Mfloat/s 188.07 ns/f
java.math.BigDecimal : 168.11 MB/s (+/- 6.9 %) 9.65 Mfloat/s 103.64 ns/f
JavaDoubleParser String : 388.09 MB/s (+/-11.3 %) 22.28 Mfloat/s 44.89 ns/f
JavaDoubleParser char[] : 454.38 MB/s (+/-11.1 %) 26.08 Mfloat/s 38.34 ns/f
JavaDoubleParser byte[] : 515.40 MB/s (+/-10.9 %) 29.58 Mfloat/s 33.80 ns/f
JsonDoubleParser String : 449.16 MB/s (+/-18.2 %) 25.78 Mfloat/s 38.79 ns/f
JsonDoubleParser char[] : 393.95 MB/s (+/-22.0 %) 22.61 Mfloat/s 44.22 ns/f
JsonDoubleParser byte[] : 525.83 MB/s (+/-18.6 %) 30.18 Mfloat/s 33.13 ns/f
JavaFloatParser String : 437.85 MB/s (+/-16.8 %) 25.13 Mfloat/s 39.79 ns/f
JavaFloatParser char[] : 469.22 MB/s (+/-11.8 %) 26.93 Mfloat/s 37.13 ns/f
JavaFloatParser byte[] : 535.85 MB/s (+/-10.0 %) 30.76 Mfloat/s 32.51 ns/f
JavaBigDecimalParser String : 481.87 MB/s (+/-19.9 %) 27.66 Mfloat/s 36.16 ns/f
JavaBigDecimalParser char[] : 573.29 MB/s (+/-18.0 %) 32.91 Mfloat/s 30.39 ns/f
JavaBigDecimalParser byte[] : 628.67 MB/s (+/-16.1 %) 36.08 Mfloat/s 27.71 ns/f
Speedup JavaDoubleParser String vs java.lang.Double : 4.43
Speedup JavaDoubleParser char[] vs java.lang.Double : 5.19
Speedup JavaDoubleParser byte[] vs java.lang.Double : 5.89
Speedup JsonDoubleParser String vs java.lang.Double : 5.13
Speedup JsonDoubleParser char[] vs java.lang.Double : 4.50
Speedup JsonDoubleParser byte[] vs java.lang.Double : 6.01
Speedup JavaFloatParser String vs java.lang.Float : 4.73
Speedup JavaFloatParser char[] vs java.lang.Float : 5.07
Speedup JavaFloatParser byte[] vs java.lang.Float : 5.78
Speedup JavaBigDecimalParser String vs java.math.BigDecimal: 2.87
Speedup JavaBigDecimalParser char[] vs java.math.BigDecimal: 3.41
Speedup JavaBigDecimalParser byte[] vs java.math.BigDecimal: 3.74
'
parsing numbers in file /Users/Shared/Developer/Java/FastDoubleParser/github/FastDoubleParser/data/canada.txt
read 111126 lines
[...]
Measurement results:
java.lang.Double : 80.77 MB/s (+/- 5.2 %) 4.64 Mfloat/s 215.44 ns/f
java.lang.Float : 96.83 MB/s (+/- 4.3 %) 5.56 Mfloat/s 179.70 ns/f
java.math.BigDecimal : 276.40 MB/s (+/- 9.5 %) 15.88 Mfloat/s 62.96 ns/f
JavaDoubleParser String : 343.75 MB/s (+/-12.1 %) 19.75 Mfloat/s 50.62 ns/f
JavaDoubleParser char[] : 440.38 MB/s (+/-11.7 %) 25.31 Mfloat/s 39.51 ns/f
JavaDoubleParser byte[] : 489.21 MB/s (+/-10.2 %) 28.11 Mfloat/s 35.57 ns/f
JsonDoubleParser String : 366.76 MB/s (+/-12.2 %) 21.08 Mfloat/s 47.45 ns/f
JsonDoubleParser char[] : 436.74 MB/s (+/-11.2 %) 25.10 Mfloat/s 39.84 ns/f
JsonDoubleParser byte[] : 512.64 MB/s (+/-10.0 %) 29.46 Mfloat/s 33.94 ns/f
JavaFloatParser String : 350.00 MB/s (+/-11.3 %) 20.11 Mfloat/s 49.72 ns/f
JavaFloatParser char[] : 476.71 MB/s (+/-13.4 %) 27.40 Mfloat/s 36.50 ns/f
JavaFloatParser byte[] : 518.69 MB/s (+/-11.3 %) 29.81 Mfloat/s 33.55 ns/f
JavaBigDecimalParser String : 386.73 MB/s (+/-14.3 %) 22.22 Mfloat/s 45.00 ns/f
JavaBigDecimalParser char[] : 569.71 MB/s (+/-16.3 %) 32.74 Mfloat/s 30.54 ns/f
JavaBigDecimalParser byte[] : 648.12 MB/s (+/-14.2 %) 37.25 Mfloat/s 26.85 ns/f
Speedup JavaDoubleParser String vs java.lang.Double : 4.26
Speedup JavaDoubleParser char[] vs java.lang.Double : 5.45
Speedup JavaDoubleParser byte[] vs java.lang.Double : 6.06
Speedup JsonDoubleParser String vs java.lang.Double : 4.54
Speedup JsonDoubleParser char[] vs java.lang.Double : 5.41
Speedup JsonDoubleParser byte[] vs java.lang.Double : 6.35
Speedup JavaFloatParser String vs java.lang.Float : 3.61
Speedup JavaFloatParser char[] vs java.lang.Float : 4.92
Speedup JavaFloatParser byte[] vs java.lang.Float : 5.36
Speedup JavaBigDecimalParser String vs java.math.BigDecimal: 1.40
Speedup JavaBigDecimalParser char[] vs java.math.BigDecimal: 2.06
Speedup JavaBigDecimalParser byte[] vs java.math.BigDecimal: 2.34
FastDoubleParser also speeds up parsing of hexadecimal float literals:
parsing numbers in file /Users/Shared/Developer/Java/FastDoubleParser/github/FastDoubleParser/data/canada_hex.txt
read 111126 lines
[...]
Measurement results:
java.lang.Double : 38.45 MB/s (+/- 5.1 %) 2.11 Mfloat/s 474.29 ns/f
java.lang.Float : 39.15 MB/s (+/- 4.2 %) 2.15 Mfloat/s 465.82 ns/f
JavaDoubleParser String : 407.92 MB/s (+/- 8.1 %) 22.37 Mfloat/s 44.71 ns/f
JavaDoubleParser char[] : 605.77 MB/s (+/- 9.0 %) 33.21 Mfloat/s 30.11 ns/f
JavaDoubleParser byte[] : 640.55 MB/s (+/- 7.7 %) 35.12 Mfloat/s 28.47 ns/f
JavaFloatParser String : 389.53 MB/s (+/- 7.0 %) 21.36 Mfloat/s 46.82 ns/f
JavaFloatParser char[] : 606.53 MB/s (+/- 7.5 %) 33.26 Mfloat/s 30.07 ns/f
JavaFloatParser byte[] : 636.46 MB/s (+/- 7.1 %) 34.90 Mfloat/s 28.66 ns/f
Speedup JavaDoubleParser String vs java.lang.Double : 10.61
Speedup JavaDoubleParser char[] vs java.lang.Double : 15.75
Speedup JavaDoubleParser byte[] vs java.lang.Double : 16.66
Speedup JavaFloatParser String vs java.lang.Float : 9.95
Speedup JavaFloatParser char[] vs java.lang.Float : 15.49
Speedup JavaFloatParser byte[] vs java.lang.Float : 16.26
Comparison of JDK versions
The Y-axis shows Mfloat/s.
| Method | 1.8.0_281 | 11.0.8 | 18.0.1.1 | 19-ea | 20-ea | 17.0.3graalvm |
|---|---|---|---|---|---|---|
| Double | 4.86 | 5.34 | 5.09 | 5.28 | 4.81 | 6.96 |
| FastDouble String | 20.30 | 27.83 | 31.18 | 32.59 | 26.15 | 32.74 |
| FastDouble char[] | 30.60 | 30.68 | 33.58 | 37.20 | 29.95 | 34.32 |
| FastDouble byte[] | 31.29 | 35.61 | 38.24 | 39.67 | 34.23 | 40.21 |
Comparison with C version
For comparison. here are the test results
of simple_fastfloat_benchmark
on the same computer:
version: Thu Mar 31 10:18:12 2022 -0400 f2082bf747eabc0873f2fdceb05f9451931b96dc
Intel(R) Core(TM) i7-8700B CPU @ 3.20GHz SIMD-256
$ ./build/benchmarks/benchmark
# parsing random numbers
available models (-m): uniform one_over_rand32 simple_uniform32 simple_int32 int_e_int simple_int64 bigint_int_dot_int big_ints
model: generate random numbers uniformly in the interval [0.0.1.0]
volume: 100000 floats
volume = 2.09808 MB
netlib : 317.31 MB/s (+/- 6.0 %) 15.12 Mfloat/s 66.12 ns/f
doubleconversion : 263.89 MB/s (+/- 4.2 %) 12.58 Mfloat/s 79.51 ns/f
strtod : 86.13 MB/s (+/- 3.7 %) 4.10 Mfloat/s 243.61 ns/f
abseil : 467.27 MB/s (+/- 9.0 %) 22.27 Mfloat/s 44.90 ns/f
fastfloat : 880.79 MB/s (+/- 6.6 %) 41.98 Mfloat/s 23.82 ns/f
OpenJDK 20-ea+22-1594
java.lang.Double : 89.59 MB/s (+/- 6.0 %) 5.14 Mfloat/s 194.44 ns/f
JavaDoubleParser String : 485.97 MB/s (+/-13.8 %) 27.90 Mfloat/s 35.85 ns/f
JavaDoubleParser char[] : 562.55 MB/s (+/-10.0 %) 32.29 Mfloat/s 30.97 ns/f
JavaDoubleParser byte[] : 644.65 MB/s (+/- 8.7 %) 37.01 Mfloat/s 27.02 ns/f
'
$ ./build/benchmarks/benchmark -f data/canada.txt
# read 111126 lines
volume = 1.93374 MB
netlib : 337.79 MB/s (+/- 5.8 %) 19.41 Mfloat/s 51.52 ns/f
doubleconversion : 254.22 MB/s (+/- 6.0 %) 14.61 Mfloat/s 68.45 ns/f
strtod : 73.33 MB/s (+/- 7.1 %) 4.21 Mfloat/s 237.31 ns/f
abseil : 411.11 MB/s (+/- 7.3 %) 23.63 Mfloat/s 42.33 ns/f
fastfloat : 741.32 MB/s (+/- 5.3 %) 42.60 Mfloat/s 23.47 ns/f
OpenJDK 20-ea+22-1594
java.lang.Double : 80.96 MB/s (+/- 5.4 %) 4.65 Mfloat/s 214.95 ns/f
JavaDoubleParser String : 415.63 MB/s (+/-10.1 %) 23.88 Mfloat/s 41.87 ns/f
JavaDoubleParser char[] : 584.31 MB/s (+/- 9.7 %) 33.58 Mfloat/s 29.78 ns/f
JavaDoubleParser byte[] : 609.54 MB/s (+/- 7.4 %) 35.03 Mfloat/s 28.55 ns/f