riptable.rt_hstack

Functions

hstack_any(itemlist[, cls, baseclass, destroy])

stack_rows or hstack_any is the main routine to row stack riptable classes.

Attributes

stack_rows

riptable.rt_hstack.hstack_any(itemlist, cls=None, baseclass=None, destroy=False, **kwargs)

stack_rows or hstack_any is the main routine to row stack riptable classes. It stacks categoricals, datasets, time objects, structs. It can now stack a dictionary of numpy arrays to return a single array and a categorical.

Parameters:

  • itemlist – a list of objects to stack for arrays, datasets, categoricals a dictionary of numpy arrays

  • cls (None. the type of class we are stacking) –

  • baseclass (the baseclass we are stacking) –

  • destroy (bool, False. Only valid for Datasets) – !! This is dangerous so make sure you do not want the data anymore in the original datasets.

Returns:

  • In the case of a list (returns a single new array, dataset, categorical or specified object.)

  • In the case of a dict (returns a single new array and a new categorical (two objects returned).)

Examples

>>> stack_rows([arange(3), arange(2)])
FastArray([0, 1, 2, 0, 1])

>>> d={'test1':arange(3), 'test2':arange(1), 'test3':arange(2)}
>>> arr, cat = stack_rows(d)
>>> Dataset({'Data':arr, 'Cat': cat})
#   Data   Cat
-   ----   -----
0      0   test1
1      1   test1
2      2   test1
3      0   test2
4      0   test3
5      1   test3

See also

np.hstack, rt.Categorical.align, rt.Dataset.concat_rows

riptable.rt_hstack.stack_rows