Convert list in tuple to numpy array?

Question

I have tuple of lists. One of these lists is a list of scores. I want to convert the list of scores to a numpy array to take advantage of the pre-built stats that scipy provides.

In this case the tuple is called 'data'

In [12]: type data[2]
-------> type(data[2])
Out[12]: <type 'list'>

In [13]: type data[2][1]
-------> type(data[2][1])
Out[13]: <type 'list'>

In [14]: type data[2][1][1]
-------> type(data[2][1][1])
Out[14]: <type 'float'>

In [15]: print data[2][1]
-------> print(data[2][1])
[16.66, 16.66, 16.66, 16.66, 5.5599999999999996, 16.699999999999999]

In [16]: print data[2][1][1]
-------> print(data[2][1][1])
16.66

Can I do this easily once I have stored the tuple?

Solution

The command numpy.asarray will turn a number of pre-set iterable containers (list, tuple, etc) into a numpy array.

Comments

[linello]

+1 for numpy.asarray it's very powerful, just a note. does it return a deep copy or simply another view of the input array?

[John Zwinck]

@linello: It copies if it has to, such as when the input is a Python built-in sequence type.

[Imago]

Nope, that does not always work, thus -1. E.g. x = np.asarray((1,2,3,4,5))

[Seth Johnson]

Haha, this question and answer are both 11 years old! @Imago On numpy 1.19.2 and python 3.10, as well as the very early version of numpy here, asarray does work for tuples: your test code gives me a numpy array. What error are you seeing?