pyEPR issue with selction of modes

[davidjhonsonn]

​

First of all, thank you very much for your tutorials in pyEPR, they saved me!

I'm a master student of Roma3, working at INFN for a design of two transmon qubits.

The code I'm using is similar to this from Daniel Cohen Hillel : 🔗 Coupling between the readout resonator and transmon - Daniel Cohen Hillel , in which he selects only 2 of the 5 modes (otherwise the simulation is to slow).

But with the same code I got an error: '[1] not in index' when I run this cell:
epra = epr.QuantumAnalysis(eprh.data_filename)
epra.analyze_all_variations(cos_trunc = 8, fock_trunc = 15);

Probably is something related to the index, because after selecting the modes 0 and 4 it doesn't find anymore the modes 1 in the array.

I tried also selecting the mode again in the analyze_all_variations but I got the error "index 4 is out of bounds for axis 0 with size 2"

[github-actions]

👏👏👏 You are awesome! Thank you for making your first issue to pyEPR ' first issue

[zlatko-minev]

Not sure from the simple message. Depends on the file.

Check out:

pyEPR/pyEPR/core_quantum_analysis.py

Line 239 in 551ec71

def __init__(

You can see wha the loading does. You can print some of the statements there or the full error stack, not sure where the error is.

For instance print
epra.n_modes
epra.variations
epra.freqs_hfss
epra.PM

[davidjhonsonn]

thank you for the code.

For instance print
epra.n_modes = 5
epra.variations = 1
epra.freqs_hfss = gives me the a panda.series with the 5 frequencies
epra.PM = gives me the a panda.series with 5 numbers

If you need I can send the full output but I don't think the problem is there. Perhaps is something related to the old version of pandas (I'm using version 1.5.3).

The strange thing is that it works if I select the modes [0, 1, 2], otherwise not. for example, If I select the modes [0,2,3] it gives me an error that index 1 is missing. Now I'm trying to change the order of the modes, [0,2,4,3,1] and then select the first three. Anyway it's not easy, if you know how to do change the order of the modes please help me. For instance, after I changed the order of modes, freqs_hfss_GHz remains uncghanged. So I manually change it. But now f_ND results are unchanged, too.

[zlatko-minev]

Yes, it could be the indexing. What's the print stack trace of the error / what is the line where this happens

[davidjhonsonn]

I solved the problem changing the code of pyEPR\core_quantum_analysis, line 670, where it gives me the error of indexing.
Let me know if my solution it's not a good one and if you have other idea.
Thank you for your attention.

I select the mode here:
modes=[1,2,4]
eprDist.modes = modes

var_select = None

data_file,var_return = eprDist.do_EPR_analysis(variations=var_select, modes=modes)

Then I have to run the quantum analysis.Here is the line:
eprQ.analyze_all_variations(**Qanalysis_settings, variations=var_select, modes=modes)

The problem is that if I don't select the modes in the quantum analysis it dosen't recognize the index (for instance with modes [0, 2, 4] it says index 1 is missing.

If I select the modes also here it says "index 4 is out of buonds for axis 0 with size 3". The problem is that the code uses the array modes to address the variables like PJ, SJ and so on. So, when the modes are [0, 2, 4] it try to access to the fifth element of PJ but it has only three elements (0, 2 and 4).

So I changed the original code using an array of indices instead of the array of the modes for those variables: