[Tutor] Setting thresholds in a compact way

[Mats Wichmann]

On 1/1/20 9:20 AM, Robert Alexander wrote:

> Not even quite sure that using floats (albeit single decimal) as dictionary
> keys will be ok.

It's not ideal - it isn't illegal, but comparison of binary floats can 
be trickier than you expect (not just in Python!), and because 
dictionary lookups are based on hashing the key, you can get interesting 
problems.

It's fairly easy to mitigate that problem - round your threshold value, 
or use decimal floats, or use the old trick of multiplying by 10 and 
calling it an int - that is, for 0.4, use 4, etc.

dictionaries, however, aren't really that natural a fit for this kind of 
thing. They're perfect when you're doing an exact match - give me the 
value for key "foo".  But you want find where you fall in ranges instead.

As others have said, I think the if/elif sequence would be fine for 
this, although it will look ugly if you have too many different buckets.