如何在C中表示科学记数法

Question

How do I represent extremely large or small numbers in C with a certain amount of significant figures. For example, if I want to do calculations on 1.54334E-34, how could I do this. Also, is this applicable to OpenCL code?

Answer 1

float var = 1.54334E-34;
double var2 = 1.54334E-34;

printf("\n normal:%f\n sci:%e \n or \n sci:%E   \n",var,var,var);
printf("\n normal:%f\n sci:%e \n or \n sci:%E   \n",var2,var2* 1.0E3 ,var2 * 1.0e3);

Answer 2

I don't know any OpenCL but 32-bit C floats will hold values in the range of +/- 3.4e +/- 38 (~7 digits) , and doubles much more. If you want arbitrary precision arithmetic/math you may want to look into GMP or MPFR .