How to efficiently copy from a double array to a byte array

Question

1.00/5 (1 vote)

See more:

I'm trying to process a double array that has over 1 million entries and tried the following routine listed below. However I experienced a very long processing time, which for example a 1000 points took over 5 minutes to process. Here is a snippet of the code I have constructed to perform this task. Is there a more efficient way speed things up when converting from a double[] to a byte[]?

double[] val;
byte[] rawDataByte = new byte[val.Rows];

for (int i = 0; i < val.Rows; i++)
{

rawDataByte[i] = (byte)val.Vector[i];
}

Thanks In Advance.

Posted 28-May-14 12:26pm

lino_76

Add a Solution

Comments

Sergey Alexandrovich Kryukov 28-May-14 19:10pm

What are you writing here? "val" does not have the member "Rows"; "val.Rows" won't compile. Please, always post some really working code.
I would first of all think why do you have this type cast. Most of your time is spend on type cast.
—SA

phil.o 28-May-14 19:45pm

Does not have a 'Vector' property, either :)

Sergey Alexandrovich Kryukov 28-May-14 20:58pm

I am trying to guess: where could it come from? Gibberish...
—SA

Lino S 28-May-14 21:14pm

Thank You Sergey and Philo for your comments. Yes you are both correct. Basically the code I have written was to work with MATLAB. I'm receiving data from MATLAB via IntPtr reference. From that reference, I am converting it to data type of double[]. I now have to convert this data to a byte array. My real question was focused more on how to efficiently convert data of type double array to byte array. The for loop listed above takes a long time when casting from double to byte. I hope this helps. Thank Again.

2 solutions

Add a Solution

Add your solution here

Treat my content as plain text, not as HTML

Preview 0

…

Existing Members

Sign in to your account

...or Join us

Download, Vote, Comment, Publish.

Your Email
Password
Forgot your password?

Your Email
This email is in use. Do you need your password?
Optional Password

I have read and agree to the Terms of Service and Privacy Policy
Please subscribe me to the CodeProject newsletters

When answering a question please:

Read the question carefully.
Understand that English isn't everyone's first language so be lenient of bad spelling and grammar.
If a question is poorly phrased then either ask for clarification, ignore it, or edit the question and fix the problem. Insults are not welcome.
Don't tell someone to read the manual. Chances are they have and don't get it. Provide an answer or move on to the next question.

Let's work to help developers, not make them feel stupid.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

phil.o · Answer 1 · 2014-05-28T13:40:00

Solution 2

As Sergey told you, the code you show won't even compile, at least in C#.

The best, most efficient method is to use the Buffer class.

C#

int count = 10;
double[] reals = new double[count];
Random r = new Random();
for (int i = 0; i < count; i++) {
   reals[i] = r.NextDouble() * 2000000000.0;
}
count <<= 3; // same as count *= 8 - because a double is 64 bits, thus 8 bytes
byte[] result = new byte[count];
Buffer.BlockCopy(reals, 0, result, 0, count);

This will copy the core bits of the double array to the byte array, in little endian order.
If you need it in big endian order, you need to:
- whether make a result.Reverse() after Buffer.BlockCopy().
- or implement your own copy function.

Buffer.BlockCopy Method[^]

Please, before posting a question, be honest and show some actually compiling code.

Posted 28-May-14 13:40pm

phil.o

Updated 28-May-14 14:14pm

v4

Comments

Sergey Alexandrovich Kryukov 28-May-14 21:01pm

Did you test it? If it was different integer types, it would work, but floating-point layout is very different. And a double is not guaranteed to fit in byte. I would think at not having those doubles in first place or not using bytes...
—SA

phil.o 29-May-14 5:15am

Never had to, but I know it is possible. It just fits the double's bit representation into a byte array.
From the MSDN page:
"Buffer only affects arrays of primitive types; this class does not apply to objects. Each primitive type is treated as a series of bytes without regard to any behavior or limitation associated with the primitive type.
Buffer provides methods to copy bytes from one array of primitive types to another array of primitive types, get a byte from an array, set a byte in an array, and obtain the length of an array. This class provides better performance for manipulating primitive types than similar methods in the System.Array class.
Buffer is applicable to the following primitive types: Boolean, Char, SByte, Byte, Int16, UInt16, Int32, UInt32, Int64, UInt64, IntPtr, UIntPtr, Single, and Double."

After all, a floating point number still has a fixed binary reprentation internally. What to do with it is a totally different story :)

Sergey Alexandrovich Kryukov 29-May-14 9:36am

That's exactly the thing: this is a matter of concrete types. You are actually doing a sort of reinterpret typecast, with shift. It may or may not work. What works on 2s-complement may not work for floating-point and 2s-complement (okay, positive bytes).
And in all cases, if double is 2.11*e4, no "conversion" to byte can work. :-)
—SA

Mohammed Nabeel Khan · Answer 2 · 2014-05-28T13:31:00

Solution 1

i would prefer Buffer.BlockCopy check this link

Posted 28-May-14 13:31pm

Mohammed Nabeel Khan