Searching for an index of max element in a "list of lists" using only specific axis-CodePudding

I am new to Python, so please help.

I need to find the max element and it's index in a two-dimensinal list, but I need to search only one column. It is better to explain using example:

Initial list of lists:

a = [[1.0, 1.0], [2.0, 2.0], [4.0, 4.0], [8.0, 1.0]]

I need to search the position x in a, where a[x][1] will be max in a[x][1]. So I don't need to search for a max in a "specified column".

The output must give x = 2, because a[2][1] == 4

I don't want to scan all the list in a cycle manually, I am searching for the elegant way in Python.

What approach should I use?

I have tried to use

max = max(a)
print(max)

but it gives a [8.0, 1.0] output - the element of a list with the maximum element inside.

OR for example,

max2 = np.max(a, axis=1)
max2_indexes = np.argmax(a, axis=1)
print(max2)
print(max2_indexes)

that outputs for provided data

[4. 2. 1. 8.] [0 0 0 0]

I am suggesting to use Numpy axis=0 parameter but maybe there is a more naive way of doing this without cycles?

max2_indexes = np.argmax(a, axis=0)

In this case I have a right answer [3 2], but maybe it's a wrong approach?

(The real list is very large - about 100000 of "small lists")

CodePudding user response：

In [2]: a = [[1.0, 1.0], [2.0, 2.0], [4.0, 4.0], [8.0, 1.0]]

A "transpose" of such a nested list is:

In [3]: a1 = list(zip(*a))    
In [4]: a1
Out[4]: [(1.0, 2.0, 4.0, 8.0), (1.0, 2.0, 4.0, 1.0)]

And from that you can easily get the max of "2nd column":

In [5]: a1[1]
Out[5]: (1.0, 2.0, 4.0, 1.0)    
In [6]: max(a1[1])
Out[6]: 4.0

or we can just get the 2nd column with a simple list comprehension:

In [13]: max([i[1] for i in a])
Out[13]: 4.0

numpy makes treating rows and columns like this easier - but at the expense of converting the list to array:

In [7]: A= np.array(a)    
In [8]: A.max(axis=0)
Out[8]: array([8., 4.])

and it makes finding the index easier:

In [9]: A.argmax(axis=0)
Out[9]: array([3, 2], dtype=int64)

You can select the 2nd column either before or after the argmax:

In [10]: A.argmax(axis=0)[1]
Out[10]: 2
In [11]: A[:,1].argmax()
Out[11]: 2

CodePudding user response：

You can get the maximum value in a nested list by using max function :

max(a,key = lambda x:x[1])

#output
[4.0, 4.0]

Now, when you have the maximum value you can find the index by using enumerate and looping:

for x,y in enumerate(a):
    if y==max(a,key = lambda x:x[1]):
        print(x)

#output
2