Hi,

I have a business dataset, which has a few columns:

 DeviceModel, DeviceSN, Bin, DurationInMin

The DurationInMin tells how long the device stay in each Bin. The range of Bin is from 1 to 20.  Normally most devices stay in the lower bin for a long time. We want to know the difference of the duration in each bin by each device model. 

What's the best way to find a metric to find the distribution of the data?  My first attempt is to calculate percentage of time each device spend in each bin, then sum it up by bin. The cons for this method is if  we have devices always stay in one bin for long time, it will blow up our data. I need help with finding other ways help with this problem.

Thanks,

 Eci