PyProcar QE parser fails for non-collinear calculations without SOC (s_z inst

[p20250462-coder]

Dear PyProcar Developers,

I am using PyProcar to analyze band projections from a Quantum ESPRESSO calculation with non-collinear magnetism enabled but without spin–orbit coupling.

QE settings:

noncolin = .true.
lspinorb = .false.

When running pyprocar.bandsplot() with the kpdos.out file, the QE parser crashes with the following error:

IndexError: list index out of range

The error originates from the following line in pyprocar/io/qe.py:

j_orbital_type_index = float(re.findall("(?<=j=)\s*([- \d.]*)", wfc)[0])

The parser assumes that the orbital description contains a j= value.

However, in non-collinear calculations without SOC, Quantum ESPRESSO prints the atomic states using s_z instead of j.

Example from kpdos.out:

The atomic states are defined as follows:
state # 1: atom 1 (Cr1), wfc 1 (l=0 m=1 s_z= 0.5)
state # 2: atom 1 (Cr1), wfc 1 (l=0 m=1 s_z=-0.5)
state # 5: atom 1 (Cr1), wfc 3 (l=1 m=1 s_z= 0.5)

Because j= does not appear in these lines, the regex search returns an empty list and causes the crash.

It seems the parser currently assumes SOC-style orbital notation, while QE outputs spin-projection notation (s_z) when lspinorb = .false..

---