I am working with many lists in which float values will be stored.
Is there a simple way to specify the type of a list when creating it?
Since I continuously add values to the list in a loop, I would like to avoid the application of np.float_()
. Can I specify that all variables that are added are automatically converted to the float type?
Thanks in advance!
Lists in standard out-of-the-box Python are not limited to a particular data type, so it is not foreseen to define one when creating the list. Numpy arrays, on the other hand, have that limitation, so I think your best bet would be to create one via np.array()
and specify the data type float
with the dtype
parameter, e.g.:
arr = np.array([7, 8, 9], dtype=float)