ProjRGD_ITE

Implementation of projected Riemannian gradient descen. As used in paper: Equating quantum imaginary time evolution, Riemannian gradient flows, and stochastic implementations (see arXiv:2504.06123)

To integrate the QITE algorithm into this software, clone the git repository into the QITE folder from https://github.com/mariomotta/QITE (make sure there is no QITE folder in the already present QITE folder). The following edits are needed to allow Julia to make use of the files:

In function QITE in file code_v4/qite.py (starting line 143) introduce the lines:

• FIDE_LIST = np.zeros(nbeta) at line 186
• FIDE_LIST[ib] = fide after fout.flush()
• return FIDE_LIST after fout.close()

For certain version of python, brackets need to placed around print statements that appear in the code.

For file code_v4/qite.py this appears on line:

• 151

For file code_v4/binary_functions.py this appears on line:

• 47

For file code_v4/pauli.py this appears on lines:

• 38
• 43

For file code_v4/hamiltonian.py this appears on lines:

• 39
• 40
• 41
• 43
• 88
• 93
• 184
• 185
• 188
• 193
• 195
• 199
• 207
• 211
• 216
• 222
• 226
• 228
• 236
• 243
• 249