Pandas provides a timeseries and stack-of-timeseries objects. They seem heavily geared towards financial data. Despite the fact of being an ndarray, Pandas objects seem to be a specialized alternative to ndarrays rather than an augmentation of them.
Features
Point-indexing syntax can use ticks or integer indices. Range indexing only works with integers, but uses the same syntax
Integer tick values interfere with integer indexing, for example:
>>> t = pandas.Series.fromValue(1.0, range(5,0,-1), 'i')
>>> t[:] = np.random.randint(100, size=5)
>>> t
5 23
4 62
3 66
2 91
1 91
>>> t[2] = 0
>>> t
5 23
4 62
3 66
2 0
1 91