Another EPR Alignment

Please do see the other EPR alignment example.

However, this EPR alignment represents a more challenging case b/c:

• We are extracting 1D slices from a 2D dataset

• The spectra are somewhat more dissimilar, and so require progressive alignment against an average

• 
• 
• 
• 
• 
• 
• 
• 

1: Raw |||mT
2: correlation |||mT
3: find center |||mT
4: before centering -- ift |||kcyc · (T)$^{-1}$
5: after centering -- ift |||kcyc · (T)$^{-1}$
6: aligned, autoscaled |||mT
7: centered spectra
8: normalized after alignment |||mT

/home/jmfranck/git_repos/pyspecdata/pyspecdata/figlist.py:782: UserWarning: This figure includes Axes that are not compatible with tight_layout, so results might be incorrect.
  plt.gcf().tight_layout()

from pyspecdata import find_file, figlist_var
from pyspecProcScripts import align_esr
import matplotlib.pyplot as plt

d = {}
d["250130 data: initial sample at 293K"] = find_file(
    "250130_20mM_TSO4_W20_Isooct_run1.DSC", exp_type="francklab_esr/romana"
)  # cooling
d["240220 data: initial sample at 293K"] = find_file(
    "240220_TempoSO4_W20_run1.DSC", exp_type="francklab_esr/romana"
)  # cooling
d["250206: Before cooling bath"] = find_file(
    "250206_20mM_TSO4_W20_C8.DSC", exp_type="francklab_esr/romana"
)  # cooling
for k, v in d.items():
    if "250206" in k:
        d[k] = v.chunk_auto("harmonic", "harmonic")["harmonic", 0]["phase", 0]
    else:
        d[k] = v.chunk_auto("harmonic", "harmonic")["harmonic", 0]["phase", 0]
        d[k].reorder("Temperature").sort("Temperature")
        d[k] = d[k]["Temperature":293]
with figlist_var() as fl:
    d = align_esr(d, fl=fl)
    fl.next("normalized after alignment")
    # If the data very dissimilar, the least-squares scaling doesn't
    # work as well.  For these, it makes more sense to use a peak
    # normalization.
    for k, v in d.items():
        extra_norm = abs(d[k].data.max() - d[k].data.min())
        d[k] /= extra_norm
        d[k].set_prop("scaling", d[k].get_prop("scaling") * extra_norm)
        fl.plot(
            d[k],
            alpha=0.5,
            label=f"{k}\n÷ {d[k].get_prop('scaling'):#0.3g}",
        )
    fl.adjust_spines("bottom")
    plt.ylabel("")
    plt.gca().set_yticks([])