CODEWITHSUNDEEP
pythonarrayslisttuples
Ben O.
Ben O.

Reputation: 79

How do I access a multi-dimensional list of tuples?

I am populating a multi-dimensional (2D) list of NamedTuple structures, like this:

from typing import NamedTuple

class Halo(NamedTuple):
    ID: int
    axis: str
    event_file: str

I declare an empty 2D list:

halo_arr = [[[] for i in range(nhalo_bins)] for j in range(nperbin)]

I then populate the list with the above individual structures in a loop:

 my_item = Halo(int(IDarr_shuf[j][i]), axis_arr[j], evt_file)
 halo_arr[nhalos[hbin]][hbin] = my_item

I am able to fill this 2D list with the tuple structure, and access individual elements:

e.g.

print(halo_arr([0][5].ID))

but I want to be able to access the entire array of IDs (or any other array of the tuple):

e.g.

print(halo_arr.ID)

so that I can put that entire 2D array as a dataset into an hdf5 file:

e.g.

fh5.create_dataset("/Halos/IDs",data=halo_arr.ID)

Yet, I get this error:

print(halo_arr.ID)

AttributeError: 'list' object has no attribute 'ID'

I'm at a loss here, and may not be using the optimal structure.

Upvotes: 0

Views: 696

Answers (1)

Alain T.
Alain T.

Reputation: 42143

Accessing an individual property inside that list of lists would require a more explicit expression of how to access the tuples depending on what you need as output. For example using list comprehensions:

[t.ID for row in halo_arr for t in row]   # if you want it flat

[[t.ID for t in row] for row in halo_arr] # if you want it 2D

The first one (flat list) is equivalent to:

result = list()
for row in halo_arr:         # Go through the first dimension (row is a list)
    for t in row:            # Go through tuples of the row
        result.append(t.ID)  # add ID to flat list

# result will be [id1, id2, id3, ...]

The second one (2D) is equivalent to:

result = list()
for row in halo_arr:         # Go through the first dimension (row is a list)
    IDlist = list()          # prepare a row of IDs
    for t in row:            # Go through tuples of the row
        IDlist.append(t.ID)  # add to row of IDs
    result.append(IDlist)    # add row of IDs to result

# result will be [ [id1, id2, id3,  ...],
#                  [id8, id9, id10, ...]
#                  ... ]

Upvotes: 1

Related Questions