FIX: Many fixes

doc/conf.py

@@ -107,6 +107,9 @@
 # A list of ignored prefixes for module index sorting.
 modindex_common_prefix = ["mne_nirs."]
 
+# TODO: Enable this and fix links
+# default_role = "py:obj"
+
 # The name of the Pygments (syntax highlighting) style to use.
 pygments_style = "sphinx"
 
@@ -151,13 +154,85 @@
     # copied from sklearn
     r"mne\.utils\.deprecated",
 }
-
+numpydoc_xref_param_type = True
+numpydoc_xref_aliases = {
+    "BaseRaw": "mne.io.Raw",
+    "dataframe": "pandas.DataFrame",
+    "DataFrame": "pandas.DataFrame",
+    "ConductorModel": "mne.bem.ConductorModel",
+    "Forward": "mne.Forward",
+    "Info": "mne.Info",
+    "Raw": "mne.io.Raw",
+    "RawBOXY": "mne.io.Raw",
+    "RawHitachi": "mne.io.Raw",
+    "RawNIRX": "mne.io.Raw",
+    "RawSNIRF": "mne.io.Raw",
+    "SourceSpaces": "mne.SourceSpaces",
+    "Transform": "mne.transforms.Transform",
+}
+numpydoc_xref_ignore = {
+    # words (some of these should be fixed)
+    "instance",
+    "of",
+    "or",
+    "None",
+    "array",
+    "like",
+    "shape",
+    "dtype",
+    "str",
+    "figure",
+    "pairs",
+    "mayavi.mlab.Figure",
+    "mayavi.core.api.Scene",
+    "color",
+    "Figure",
+    "String",
+    "default",
+    "colormap",
+    "length",
+    "Bool",
+    "Axes",
+    "Pandas",
+    "model",
+    "optional",
+    "output",
+    "Nilearn",
+    "in",
+    "specified",
+    "as",
+    "matplotlib.Figure",
+    "MNE",
+    "As",
+    "two",
+    "values",
+    "ResultsGLM",
+    "Contrast",
+    "p",
+    "q",
+    "Number",
+    "Array",
+    "number",
+    "n_windows",
+    "n_nirs",
+    "n_add_reg",
+    "n_frames",
+    "n_onsets",
+    "derivative",
+    "integers",
+    "lists",
+}
 
 # sphinxcontrib-bibtex
 bibtex_bibfiles = ["./references.bib", "./references-nirs.bib"]
 bibtex_style = "unsrt"
 bibtex_footbibliography_header = ""
 
+nitpick_ignore_regex = [
+    # Type hints for undocumented types
+    ("py:.*", r"mne\.io\..*\.Raw.*"),  # RawEDF etc.
+]
+
 
 # -- Options for HTML output ----------------------------------------------
 
@@ -281,6 +356,7 @@
     "numpy": ("https://numpy.org/devdocs", None),
     "scipy": ("https://scipy.github.io/devdocs", None),
     "matplotlib": ("https://matplotlib.org/stable", None),
+    "pandas": ("https://pandas.pydata.org/pandas-docs/stable", None),
     "mne": ("https://mne.tools/stable", None),
     "nilearn": ("http://nilearn.github.io/stable", None),
     "sklearn": ("https://scikit-learn.org/stable", None),

environment.yml

@@ -4,44 +4,7 @@ channels:
 dependencies:
 - python>=3.8
 - pip
-- numpy
-- scipy
-- matplotlib
-- tqdm
-- pooch>=1.5
-- decorator
-- h5io
-- packaging
-- numba
-- pandas
-- xlrd
-- scikit-learn
-- h5py
-- jinja2
-- pillow
-- statsmodels
-- jupyter
-- joblib
-- psutil
-- numexpr
-- imageio
-- spyder-kernels>=1.10.0
-- imageio-ffmpeg>=0.4.1
-- vtk>=9.0.1
-- traitlets
-- pyvista>=0.32,!=0.35.2,!=0.38.0,!=0.38.1,!=0.38.2,!=0.38.3,!=0.38.4,!=0.38.5,!=0.38.6
-- pyvistaqt>=0.4
-- qdarkstyle
-- darkdetect
-- dipy
-- nibabel
+- mne
 - nilearn
-- python-picard
-- pyqt!=5.15.3
-- mffpy>=0.5.7
-- ipywidgets
-- ipyvtklink
-- mne-qt-browser
-- pymatreader
-- pip:
-  - git+https://github.com/mne-tools/mne-python.git@main
+- statsmodels
+- pytables

examples/general/plot_11_hrf_measured.py

@@ -19,11 +19,11 @@
 viewed in the
 :ref:`waveform analysis example <tut-fnirs-processing>`.
 See
-`Luke et al (2021) <https://www.spiedigitallibrary.org/journals/neurophotonics/volume-8/issue-2/025008/Analysis-methods-for-measuring-passive-auditory-fNIRS-responses-generated-by/10.1117/1.NPh.8.2.025008.short>`_
+`Luke et al (2021) <https://www.spiedigitallibrary.org/journals/neurophotonics/volume-8/issue-2/025008/Analysis-methods-for-measuring-passive-auditory-fNIRS-responses-generated-by/10.1117/1.NPh.8.2.025008.short>`__
 for a comparison of the epoching and GLM approaches.
 
 This GLM analysis is a wrapper over the excellent
-`Nilearn GLM <http://nilearn.github.io/modules/reference.html#module-nilearn.glm>`_.
+`Nilearn GLM <http://nilearn.github.io/modules/reference.html#module-nilearn.glm>`__.
 """
 # sphinx_gallery_thumbnail_number = 9
 
@@ -58,8 +58,7 @@
 #
 #    Optodes were placed over the motor cortex using the standard NIRX motor
 #    montage, but with 8 short channels added (see their web page for details).
-#    To view the sensor locations run
-#    `raw_intensity.plot_sensors()`.
+#    To view the sensor locations run ``raw_intensity.plot_sensors()``.
 #    A sound was presented to indicate which hand the participant should tap.
 #    Participants tapped their thumb to their fingers for 5s.
 #    Conditions were presented in a random order with a randomised inter
@@ -149,7 +148,7 @@
 #
 #    For further discussion on design matrices see
 #    the Nilearn examples. Specifically the
-#    `first level model example <http://nilearn.github.io/auto_examples/04_glm_first_level/plot_first_level_details.html>`_.
+#    `first level model example <http://nilearn.github.io/auto_examples/04_glm_first_level/plot_first_level_details.html>`__.
 #
 # Next we create a model to fit our data to.
 # The model consists of various components to model different things we assume
@@ -160,7 +159,7 @@
 # (as described above).
 # We also include a cosine drift model with components up to the high pass
 # parameter value. See the nilearn documentation for recommendations on setting
-# these values. In short, they suggest `"The cutoff period (1/high_pass) should be
+# these values. In short, they suggest "The cutoff period (1/high_pass) should be
 # set as the longest period between two trials of the same condition multiplied by 2.
 # For instance, if the longest period is 32s, the high_pass frequency shall be
 # 1/64 Hz ~ 0.016 Hz".
@@ -425,7 +424,7 @@
 # -----------------
 #
 # We can also define a contrast as described in
-# `Nilearn docs <http://nilearn.github.io/auto_examples/04_glm_first_level/plot_localizer_surface_analysis.html>`_
+# `Nilearn docs <http://nilearn.github.io/auto_examples/04_glm_first_level/plot_localizer_surface_analysis.html>`__
 # and plot it.
 # Here we contrast the response to tapping on the left hand with the response
 # from tapping on the right hand.

examples/general/plot_15_waveform.py

@@ -162,10 +162,10 @@
 # remove this. A high pass filter is also included to remove slow drifts
 # in the data.
 
-fig = raw_haemo.plot_psd(average=True, amplitude=False)
+fig = raw_haemo.compute_psd().plot(average=True, amplitude=False)
 fig.suptitle("Before filtering", weight="bold", size="x-large")
 raw_haemo = raw_haemo.filter(0.05, 0.7, h_trans_bandwidth=0.2, l_trans_bandwidth=0.02)
-fig = raw_haemo.plot_psd(average=True, amplitude=False)
+fig = raw_haemo.compute_psd().plot(average=True, amplitude=False)
 fig.suptitle("After filtering", weight="bold", size="x-large")
 
 # %%

examples/general/plot_30_frequency.py

@@ -87,8 +87,8 @@
 # Rescale to be in expected units of uM.
 hrf = raw_haemo.copy().pick(picks=[0])
 hrf._data[0] = 1e-6 * (design_matrix["Tapping_Left"] + design_matrix["Tapping_Right"]).T
-hrf.pick(picks="hbo").plot_psd(
-    average=True, fmax=2, xscale="log", color="r", show=False, amplitude=False
+hrf.pick(picks="hbo").compute_psd(fmax=2).plot(
+    average=True, xscale="log", color="r", show=False, amplitude=False
 )
 
 
@@ -111,8 +111,8 @@
 
 # rescale data to fit in plot. TODO: fix this
 raw_haemo._data = raw_haemo._data * 1e-2
-raw_haemo.pick(picks="hbo").plot_psd(
-    average=True, fmax=2, xscale="log", amplitude=False
+raw_haemo.pick(picks="hbo").compute_psd(fmax=2).plot(
+    average=True, xscale="log", amplitude=False
 )
 
 
@@ -145,8 +145,8 @@
     verbose=True,
 )
 
-epochs.pick(picks="hbo").plot_psd(
-    average=True, fmax=2, color="g", xscale="log", amplitude=False
+epochs.pick(picks="hbo").compute_psd(fmax=2).plot(
+    average=True, color="g", xscale="log", amplitude=False
 )
 
 
@@ -198,14 +198,16 @@
 # removes these unwanted components and they are not visible in the
 # epoched data.
 
-fig = hrf.pick(picks="hbo").plot_psd(
-    average=True, fmax=2, color="r", show=False, amplitude=False
+fig = (
+    hrf.pick(picks="hbo")
+    .compute_psd(fmax=2)
+    .plot(average=True, color="r", show=False, amplitude=False)
 )
-raw_haemo.pick(picks="hbo").plot_psd(
-    average=True, fmax=2, ax=fig.axes, show=False, amplitude=False
+raw_haemo.pick(picks="hbo").compute_psd(fmax=2).plot(
+    average=True, axes=fig.axes, show=False, amplitude=False
 )
-epochs.pick(picks="hbo").plot_psd(
-    average=True, fmax=2, ax=fig.axes, show=False, color="g", amplitude=False
+epochs.pick(picks="hbo").compute_psd(fmax=2).plot(
+    average=True, axes=fig.axes, show=False, color="g", amplitude=False
 )
 mne.viz.plot_filter(
     filter_params,
@@ -237,12 +239,12 @@
 # The green line illustrates the signal before filtering, and the red line
 # shows the signal after filtering.
 
-fig = raw_haemo.plot_psd(
-    average=True, fmax=2, xscale="log", color="r", show=False, amplitude=False
+fig = raw_haemo.compute_psd(fmax=2).plot(
+    average=True, xscale="log", color="r", show=False, amplitude=False
 )
 raw_haemo = raw_haemo.filter(l_freq=None, h_freq=0.4, h_trans_bandwidth=0.2)
-raw_haemo.plot_psd(
-    average=True, fmax=2, xscale="log", ax=fig.axes, color="g", amplitude=False
+raw_haemo.compute_psd(fmax=2).plot(
+    average=True, xscale="log", axes=fig.axes, color="g", amplitude=False
 )
 
 
@@ -273,10 +275,9 @@
                     isi_max=max_isi,
                 )
                 raw._data[0] = raw._data[0] - np.mean(raw._data[0])
-                raw.pick(picks="hbo").plot_psd(
+                raw.pick(picks="hbo").compute_psd(fmax=2).plot(
                     average=True,
-                    fmax=2,
-                    ax=axes[rep, column],
+                    axes=axes[rep, column],
                     show=False,
                     color=sm.cmap(sm.norm(max_isi)),
                     xscale="log",

examples/general/plot_99_bad.py

@@ -115,10 +115,10 @@
 # Do not do this!!
 #
 
-fig = raw_haemo.plot_psd(average=True, amplitude=False)
+fig = raw_haemo.compute_psd().plot(average=True, amplitude=False)
 fig.suptitle("Before filtering", weight="bold", size="x-large")
 raw_haemo = raw_haemo.filter(0.05, 0.1, h_trans_bandwidth=0.2, l_trans_bandwidth=0.02)
-fig = raw_haemo.plot_psd(average=True, amplitude=False)
+fig = raw_haemo.compute_psd().plot(average=True, amplitude=False)
 fig.suptitle("After filtering", weight="bold", size="x-large")
 
 

mne_nirs/statistics/_glm_level_first.py

@@ -4,6 +4,7 @@
 
 import warnings
 from copy import deepcopy
+from inspect import getfullargspec
 from pathlib import PosixPath
 
 import numpy as np
@@ -784,8 +785,13 @@ def run_glm(raw, design_matrix, noise_model="ar1", bins=0, n_jobs=1, verbose=0):
 def _compute_contrast(glm_est, contrast, contrast_type=None):
     from nilearn.glm.contrasts import compute_contrast as _cc
 
+    key = "contrast_type"
+    _ccc = _cc
+    if hasattr(_cc, "__wrapped__"):
+        _ccc = _cc.__wrapped__
+    key = "stat_type" if "stat_type" in getfullargspec(_ccc).args else "contrast_type"
     return _cc(
-        np.array(list(glm_est.keys())), glm_est, contrast, contrast_type=contrast_type
+        np.array(list(glm_est.keys())), glm_est, contrast, **{key: contrast_type}
     )
 
 

mne_nirs/utils/_io.py

@@ -66,14 +66,18 @@ def glm_to_tidy(info, statistic, design_matrix, wide=True, order=None):
 
 def _tidy_Contrast(data, glm_est, design_matrix):
     df = pd.DataFrame()
+    if hasattr(glm_est, "stat_type"):
+        contrast_type = glm_est.stat_type
+    else:  # nilearn < 0.10
+        contrast_type = glm_est.contrast_type
     for idx, ch in enumerate(data.ch_names):
         df = pd.concat(
             [
                 df,
                 pd.DataFrame(
                     {
                         "ch_name": ch,
-                        "ContrastType": glm_est.contrast_type,
+                        "ContrastType": contrast_type,
                         "variable": "effect",
                         "value": glm_est.effect[0][idx],
                     },
@@ -82,7 +86,7 @@ def _tidy_Contrast(data, glm_est, design_matrix):
                 pd.DataFrame(
                     {
                         "ch_name": ch,
-                        "ContrastType": glm_est.contrast_type,
+                        "ContrastType": contrast_type,
                         "variable": "p_value",
                         "value": glm_est.p_value()[idx],
                     },
@@ -91,7 +95,7 @@ def _tidy_Contrast(data, glm_est, design_matrix):
                 pd.DataFrame(
                     {
                         "ch_name": ch,
-                        "ContrastType": glm_est.contrast_type,
+                        "ContrastType": contrast_type,
                         "variable": "stat",
                         "value": glm_est.stat()[idx],
                     },
@@ -100,7 +104,7 @@ def _tidy_Contrast(data, glm_est, design_matrix):
                 pd.DataFrame(
                     {
                         "ch_name": ch,
-                        "ContrastType": glm_est.contrast_type,
+                        "ContrastType": contrast_type,
                         "variable": "z_score",
                         "value": glm_est.z_score()[idx],
                     },

requirements_testing.txt

@@ -19,3 +19,4 @@ seaborn
 fooof
 snirf
 statsmodels
+tables