I hope you realize that unless your GPS is accurate to 3 meters, you are not going to actually have 3 meters of accuracy, regardless of how many decimal points it reports.

Regarding your storage issue. Single precision floats have at most 6 to 9 decimal digits of precision. When converting to a string, you are essentially generating a decimal representation of that float. This means that you are guaranteed that a number with 6 decimal digits can be converted back and forth between a float and it’s string representation (which is decimal). However, any more than that really depends on the number (ie some 7, 8, and 9 digit decimals can be converted exactly while others won’t).

So, in your case, you have an 8 digit decimal number that you first convert to float, then back to decimal (in string form). You should expect that at most, the first 6 digits will match.

In you second example, you used atoi which converts an ascii string to an integer. Integers are whole numbers. They don’t have decimal points. So, the output is exactly what you should have gotten. Try atof.

Any ideas? Normal scientific notation (e.g. 5.1498587E+03 or 1.2039404-03 etc is no good, I can’t use if for what I need to do)

Is there any way to print the value as well as parsing the float into a string while keeping the precision?

float f2= -80.123456;
sprintf(op, “string %3.6f”, f2);
printf(op);

returns: string -80.123459

– OR –

Is there another way I can parse the value from a string to a decimal and keep the precision?

char read[10] = “-80.123456″;
float lon = (1000000 * (float)atoi(read));
printf(“lf %3.6f\n”, lon);

returns: lf -80000000.000000

I am willing to split the char value into 3 integers, (high=”-80″, mid=”123″, low=”456″) but not sure how to parse it in to parts while maintaining precision. (value range “123.123456″ to “-101.123456″ read as a string)