Instead of only being able to get the integral and its integration time,
it can also be handy to get the integral divided/differentiated by the
the integration time. This data is then just filtered by the integrator.
This adds a second put method to the integrator class. This allows to
integrate with known intervals between samples, rather than based on
timestamps. This makes integrating the samples coming out of the MPU9250
FIFO buffer easier.
The integrator had an untested read mode which did not apply the coning
correction. Instead of keeping two integrals (auto/read) it is now one
and the reset mechanism can be selected by setting the
auto_reset_interval to 0 to disable it or some positive number else.
Also, the integrator could potentially explode if a (single) timestamp
was wrong, so before the last integrated one. This is now caught with a
dt of 0 instead of inf/nan.
