Matrix user forums

Hello Benj,All,
I would like, for reasons of precision, to be able to memorize in the EEPROM for example a value U = 4.9542
I tried by creating a 16-bit shift variable, with a variable U as Integer
MSB = U >> 8
LSB = U
=> EEVAL = (MSB << + LSB
eeprom1::write(0,MSB)
eeprom1::write(1,LSB)
It works but on 10bit only and the accuracy is not enough.
I can not either multiply 4.9542 * 10000 and recreate the shift variable, because the variable is type of Float and I would not get as a result 4.9542 * 10000 = 40000 (if I'm not mistaken).
Please, how can I do? Do you have an example ?
Thank you in advance for your replies

Hello,

You could multiply out the value first using floating point maths.

4.9542 * 10000 = 40000

This is correct. However...

4.9542 * 10000.0 = 49542.0

Which you can then copy to your integer variable.

Other ways include converting the float to a string and storing the string or splitting the float up into an array of 4 bytes using a union and then recombining directly into a float. The latter method is C only and won't simulate.

Thank you Benj,
I will try the first solution.
As for the second that you propose, I tried in C but I was not successful, during the compilation I got a lot of errors. Do you have an example (under the elbow).
Luc

Excellent answer by Benj. Thank you.

I agree with ruchikadmore..Thank you!

In Flowcode 8.1 there is a new component available under Data -> Storage called Type Conversions.

This allows you to convert between different types of variable e.g. a float can be converted into 2 unsigned ints or 4 bytes and visa versa.