Interactive Apodization

import matplotlib.pyplot as plt
import matplotlib_inline.backend_inline

matplotlib_inline.backend_inline.set_matplotlib_formats("retina")
plt.rcParams["figure.dpi"] = 150

Apodization (windowing) lets you trade off between SNR and spectral resolution when processing FID signals. Choosing the right line broadening (LB) or Gaussian broadening (GB) parameters typically requires visual trial and error.

The xmris apodization widget lets you adjust these parameters with sliders and immediately see the effect on both the time and frequency domains.

For background on the underlying filter functions and their mathematical definitions, see the Apodization processing guide.

import matplotlib.pyplot as plt
import xarray as xr

import xmris
from xmris import simulate_fid

1. Generate Synthetic Data¶

We create a noisy FID with two closely spaced peaks to demonstrate resolution enhancement.

da_fid = simulate_fid(
    amplitudes=[10.0, 8.0],
    frequencies=[150.0, 165.0],
    dampings=[15.0, 15.0],
    reference_frequency=298.0,
    carrier_ppm=4.7,
    spectral_width=2000.0,
    n_points=2048,
    target_snr=15.0,
)

2. Launch the Widget¶

Pass a 1D time-domain DataArray to the apodize method. The widget computes the Fourier transform in the browser and renders the FID (top) and spectrum (bottom) side by side.

da_fid.xmr.widget.apodize(
    unit="hz",
    lb_range=(0, 20.0),
    gb_range=(0, 30.0),
)

from xmris.visualization.widget._static_exporter import export_widget_static
from xmris.visualization.widget.apodizer.apodizer import apodize_interactive

export_widget_static(
    apodize_interactive,
    da_fid,
    unit="hz",
    lb_range=(0.0, 20.0),
    gb_range=(0.0, 30.0),
    width=740,
    height=550,
)

Widget Controls¶

3. Extract and Apply Parameters¶

Once you are satisfied with the result, click Close. The widget displays a copyable Python snippet so your parameters are recorded in your notebook.

Paste the generated code into the next cell:

# Pasted directly from the widget completion screen
da_apodized = da_fid.xmr.apodize_lg(lb=5.20, gb=6.00)

fig, ax = plt.subplots(figsize=(8, 3))
da_fid.xmr.to_spectrum().real.plot(ax=ax, color="gray", label="Original")
da_apodized.xmr.to_spectrum().real.plot(ax=ax, color="tab:blue", label="Apodized")
ax.set_title("Spectrum Processed via Widget Parameters")
ax.set_xlim(100, 220)
ax.legend()
plt.show()

assert "apodization_lb" in da_apodized.attrs
assert "apodization_gb" in da_apodized.attrs
assert da_apodized.attrs["apodization_lb"] == 5.20
assert da_apodized.attrs["apodization_gb"] == 6.00