Electron occupation bug

[ImJudgement7497]

For the simple particle in a box system with no interaction, iDEA.methods.non_interacting.solve() gives a single body state with occupancy of the up and down electrons at a given k value.

However at some multiplet states of this system (for example the 6th and 7th excited state), the occupancy remains the same between states, and does not exchange spin configurations with each other (i.e. state 6 has orbital configuration for up, down respectively ([1], [2]) but state 7 has ([1], [2]) not ([2], [1]).

Running the below see_bug.py script and looking at the resulting bug_results.txt will show what states have the same occupancy, where they should be exchanged as described.

import iDEA
import numpy as np

# iDEA initalisation 
l = 5
points = 100
x = np.linspace(0, l, points)
v_ext = np.zeros(len(x))
v_int = np.zeros([len(x), len(x)])
pib_double = iDEA.system.System(x, v_ext, v_int, electrons = "ud")
occupied_info = []
string_lines = []

for i in range(30):
    t = iDEA.methods.non_interacting.solve(pib_double, k=i)
    up_occ_index = t.up.occupied # index of occupied orbitals
    down_occ_index = t.down.occupied
    occupied_info.append([up_occ_index, down_occ_index])

for i in range(len(occupied_info)):

    if occupied_info[i] == occupied_info[i-1]:
        string_lines.append(f"State {i-1} and {i} have the same occupancy")

with open("bug_results.txt", 'w') as file:
    for line in string_lines:
        file.write(line + '\n')

Thank you!

[ImJudgement7497]

checking this request: https://github.com/iDEA-org/iDEA/pull/16/files

it seems like when doing a local install from a cloned repository, the changes in the pull request went through fine.
However when doing an install from pip, the changes do not go through, and this is why the bug happens.