ctmm-initiative / ctmmweb

Web app for analyzing animal tracking data, built upon ctmm R package
http://biology.umd.edu/movement.html
GNU General Public License v3.0
32 stars 21 forks source link

how many digits should we preserve? #57

Closed xhdong-umd closed 6 years ago

xhdong-umd commented 6 years ago

We have a lot of numbers in different places, with different meaning and units.

I'm thinking maybe we should have a consistent format, like always keep 2 digits after decimal place (3 digits will be a little bit long in some places).

@jmcalabrese @chfleming @NoonanM What do you think?

xhdong-umd commented 6 years ago

In second thought, some places probably need more precision.

For example the overlap values are 0 - 1, the current 4 digits seemed to be fine. And 3 digits is good for the model summary table DOF etc.

chfleming commented 6 years ago

For numbers with confidence intervals, its the confidence interval that determines this. In general, I think CIs need at least 1-2 significant digits on their difference, like x = 1.2345 (1.2340 -- 1.2350) for example.

xhdong-umd commented 6 years ago

I assume I need to calculate this after getting calculation result from ctmm?

chfleming commented 6 years ago

Yes.

xhdong-umd commented 6 years ago

Now the model summary table will make sure the CI values have at leasts one significant digit on difference. Previously the last rows in table below were rounded to 2 digits and didn't show the difference.

screen shot 2018-03-27 at 8 59 07 pm
xhdong-umd commented 6 years ago

Overlap table are also adjusted to reflect the significant digits needed for the difference.

screen shot 2018-03-28 at 8 56 54 am