fix typos using codespell

docs/whats_new.rst

@@ -41,7 +41,7 @@ Changelog
 - Changed RANSAC so that "bad by high-frequency noise" channels are retained when making channel predictions (provided they aren't flagged as bad by any other metric), matching MATLAB PREP behaviour, by `Austin Hurst`_ (:gh:`64`)
 - Added a new flag ``matlab_strict`` to :class:`~pyprep.PrepPipeline`, :class:`~pyprep.Reference`, :class:`~pyprep.NoisyChannels`, and :func:`~pyprep.ransac.find_bad_by_ransac` for optionally matching MATLAB PREP's internal math as closely as possible, overriding areas where PyPREP attempts to improve on the original, by `Austin Hurst`_ (:gh:`70`)
 - Added a ``matlab_strict`` method for high-pass trend removal, exactly matching MATLAB PREP's values if ``matlab_strict`` is enabled, by `Austin Hurst`_ (:gh:`71`)
-- Added a window-wise implementaion of RANSAC and made it the default method, reducing the typical RAM demands of robust re-referencing considerably, by `Austin Hurst`_ (:gh:`66`)
+- Added a window-wise implementation of RANSAC and made it the default method, reducing the typical RAM demands of robust re-referencing considerably, by `Austin Hurst`_ (:gh:`66`)
 - Added `max_chunk_size` parameter for specifying the maximum chunk size to use for channel-wise RANSAC, allowing more control over PyPREP RAM usage, by `Austin Hurst`_ (:gh:`66`)
 - Changed :class:`~pyprep.Reference` to exclude "bad-by-SNR" channels from initial average referencing, matching MATLAB PREP behaviour, by `Austin Hurst`_ (:gh:`78`)
 - Changed :class:`~pyprep.Reference` to only flag "unusable" channels (bad by flat, NaNs, or low SNR) from the first pass of noisy detection for permanent exclusion from the reference signal, matching MATLAB PREP behaviour, by `Austin Hurst`_ (:gh:`78`)

examples/run_full_prep.py

@@ -120,7 +120,7 @@
 #
 # To validate each step of pyprep's results, we compare results after each step
 # with the results from EEGLAB's PREP. To make it easy to compare, we rescale
-# the EEG data to [-1, 1] (devided the data by maximum absolute value) when
+# the EEG data to [-1, 1] (divided the data by maximum absolute value) when
 # making the plot.
 
 EEG_raw = raw_copy.get_data(picks="eeg") * 1e6
@@ -319,7 +319,7 @@
 #
 # It can be seen the results match well on each step except the final step.
 # This is due to the difference of find_noisy_channel functions, since the
-# channels with relatively large error corrsponds to the channels that are only
+# channels with relatively large error corresponds to the channels that are only
 # interpolated in Python or Matlab.
 #
 # We think the differences mainly arise from

pyprep/prep_pipeline.py

@@ -75,10 +75,10 @@ class PrepPipeline:
     raw : mne.io.Raw
         The data including eeg and non eeg channels. It is unprocessed if
         accessed before the fit method, processed if accessed after a
-        succesful fit method.
+        successful fit method.
     raw_eeg : mne.io.Raw
         The only-eeg part of the data. It is unprocessed if accessed before
-        the fit method, processed if accessed after a succesful fit method.
+        the fit method, processed if accessed after a successful fit method.
     raw_non_eeg : {mne.io.Raw, None}
         The non-eeg part of the data. It is not processed when calling
         the fit method. If the input was only EEG it will be None.

tests/conftest.py

@@ -46,7 +46,7 @@ def raw_clean(montage):
     This fixture downloads and reads in subject 30, run 2 from the Physionet
     BCI2000 (eegbci) open dataset, which contains no bad channels on an initial
     pass of :class:`pyprep.NoisyChannels`. Intended for use with tests where
-    channels are made artifically bad.
+    channels are made artificially bad.
 
     File attributes:
     - Channels: 64 EEG