Floating point analog input values aren't decoded correctly

[amcewen]

The analog_input value type will include up to two decimal places of any floating-point value that's stored in it. Currently the decoder only returns integers, and so throws away the decimal part of any value.

For example, if I encode (in micropython on a pycom board:

lpp = CayenneLPP()
lpp.add_analog_input(1, 123.456)
lpp.add_analog_input(2, -123.456)

Then the resulting buffer is '010230390202cfc6'

Passing that into decode I get:

>>> decode('010230390202cfc6')
[{'value': 123, 'name': 'Analog Input', 'channel': 1}, {'value': -124, 'name': 'Analog Input', 'channel': 2}]

Changing the 100 in analog_input_output in methods.py to 100.0 seems to fix things. (Presumably there'll be similar issues in other methods like temperature, humidty, etc., but I've not tested those)

(It's not causing any issues for me, but I figured I should report it so you're aware of it)

[OlegZv]

@amcewen Thanks a lot for reporting this! Good catch

[OlegZv]

I appreciate the report @amcewen once again.

Changes have been pushed to the master branch