pint's usage with numpy's concatenate

Question

I have this toy example where I am using pint and numpy -

import numpy as np

from pint import UnitRegistry

ureg = UnitRegistry()

Q_ = ureg.Quantity

args = []

for k in range(0,17):
   uwnd = Q_(np.ones((1,73,144)),'meter / second')
    args.append(uwnd)
uwndTot = np.vstack(args)
print(uwndTot.shape)
for element in args:
   print(type(element))

The shape of uwndTot is

 (17,73,144)

As shown above the type of the element in args is

<class 'pint.quantity.build_quantity_class.<locals>.Quantity'>

But instead if I print out the type of uwndTot in the following way I get

 for element in uwndTot:
    print(type(element))

I get

 <class 'numpy.ndarray'>

So I am unable to extract the units of uwnd from this procedure. Is that correct behavior ? Why does the type change when I issue the call

  np.vstack(args)

or

 np.concatenate(args,axis = 0)

I want to be able to use vstack or concatenate. Is there another way ?

UPDATE I have a bunch of netCDF files which when I read in are of shape (73,144). I need to concatenate 'n' of them which will give me a numpy array of shape (17,73,144). Those netCDF files have units of meters / second or other units. Those units are required in my calculation.

Answer 1

Here's what we use in MetPy to do this, making sure that all source arrays have compatible dimensionality:

def concatenate(arrs, axis=0):
    dest = 'dimensionless'
    for a in arrs:
        if hasattr(a, 'units'):
            dest = a.units
            break

    data = []
    for a in arrs:
        if hasattr(a, 'to'):
            a = a.to(dest).magnitude
        data.append(np.atleast_1d(a))

    return units.Quantity(np.concatenate(data, axis=axis), dest)