Python: Get the maximum of a certain index in a dictionary of arrays?

I have a dictionary set up like this:

{attribute: [threshold, infogain]}

where attribute is a string and infogain is a number. The list is always in the same format, so infogain’s index is always 1.

and I would like to find the best attribute based on the value of infogain. I do need to keep threshold in there, or at least I need a way to store this so that all three values are associated and accessible.

My current attempt to get the max value of infogain is this:

max(attribute.values()[1])

Which causes an understandable error saying I’m trying to subscript dict_values.

and I was thinking about how to modify max(attributes, key=attributes.get) to somehow get me the dictionary key for the array with the max infogain, but no luck.

How would I be able to do this?

>Solution :

You can take the max over the key/value pairs of the dictionary:

max_kv = max(d.items(), key = lambda kv:kv[1][1])

Max key is:

max_kv[0]  # since max_kv is a key/value pair

Example:

d = {
    "a1":[0.9, 1],
    "a2":[.25, 2],
    "a3":[.5, 5]
}
max_kv = ('a3', [0.5, 5])

So, max key is:

max_kv[0], which is a3