How do i get Matlab to recognize an input file in two's complement?

Hey,

I use MATLAB 7.9.0 (R2009b).

I'm trying to input a file generated from a modelsim simulation, and I want matlab to identify the data as two's complement, and then convert it to decimal. For instence - 0100.010100 would be converted to 4.3125.

The file contents:

0100.010100

0010.001010

0001.000101

0000.100011

0000.010001

0000.001001

0000.000100

0000.000010

0101.001100

0010.100110

0110.011110

1000.011010

1001.011000

1001.110111

1010.000110

1010.001110

1010.010010

0101.001001

0010.100101

0001.010011

The best I could do is get Matlab to identify it as decimal, and then manually write a function which converts the numbers to decimal according to two's complement rules. Got a few bugs there, and it doesn't make sence that I need to go that way...

Is there any easier way to do this.

BTW, I can manipulate the manner in which the data is written in the file (from the simulation). It can be decimal (although the RHS won't be considered two's decimal when writing to file), I can put a space instead of "." Almost any manipulation possible - except writing in true decimal value.

Thanks!

-Josh

Answer 1

[Walter Roberson]

Use no delimiter at all, just 10 binary digits in a row. Read the lines as strings. Use bin2dec() on the strings. Divide the result by 2^6. Any result that is 8 or greater, subtract that result from 8 to get the negative number being represented.