no more need for FeaturePredecessor with empty (default) value

Author: avivdotan

notebooks/scratch_features.ipynb

@@ -65,7 +65,8 @@ def __init__(
     def _validate_execution_tree(self, feature_extractors):
         for fname, f in feature_extractors.items():
             f = _get_underlying_func(f)
-            assert type(self) in f.parent_extractor_type
+            pe_type = getattr(f, "parent_extractor_type", [FeatureExtractor])
+            assert type(self) in pe_type
         return feature_extractors
 
     def _check_is_fitable(self, feature_extractors):

notebooks/scratch_features2.ipynb

@@ -29,7 +29,6 @@ def _channel_app_samp_entropy_counts(x, m, r, l):
     return kdtree.query_radius(x_emb, r, count_only=True)
 
 
-@FeaturePredecessor()
 class EntropyFeatureExtractor(FeatureExtractor):
     def preprocess(self, x, m=2, r=0.2, l=1):
         rr = r * x.std(axis=-1)
@@ -57,7 +56,6 @@ def complexity_sample_entropy(counts_m, counts_mp1):
     return -np.log(A / B)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def complexity_svd_entropy(x, m=10, tau=1):
     x_emb = np.empty((*x.shape[:-1], (x.shape[-1] - m + 1) // tau, m))
@@ -68,7 +66,6 @@ def complexity_svd_entropy(x, m=10, tau=1):
     return -np.sum(s * np.log(s), axis=-1)
 
 
-@FeaturePredecessor()
 @univariate_feature
 @nb.njit(cache=True, fastmath=True)
 def complexity_lempel_ziv(x, threshold=None):

src/eegdash/features/extractors.py

@@ -1,6 +1,6 @@
 from itertools import chain
 import numpy as np
-from scipy.signal import csd, coherence
+from scipy.signal import csd
 
 from ..extractors import FeatureExtractor, BivariateFeature
 from ..decorators import FeaturePredecessor, bivariate_feature
@@ -14,7 +14,6 @@
 ]
 
 
-@FeaturePredecessor()
 class CoherenceFeatureExtractor(FeatureExtractor):
     def preprocess(self, x, **kwargs):
         f_min = kwargs.pop("f_min") if "f_min" in kwargs else None

src/eegdash/features/feature_bank/complexity.py

@@ -4,7 +4,7 @@
 import scipy.linalg
 
 from ..extractors import FitableFeature
-from ..decorators import FeaturePredecessor, multivariate_feature
+from ..decorators import multivariate_feature
 
 
 __all__ = [
@@ -22,7 +22,6 @@ def _update_mean_cov(count, mean, cov, x_count, x_mean, x_cov):
     cov[:] -= np.outer(mean, mean)
 
 
-@FeaturePredecessor()
 @multivariate_feature
 class CommonSpatialPattern(FitableFeature):
     def __init__(self):
@@ -77,6 +76,7 @@ def fit(self):
         for l in range(len(self._labels)):
             self._covs[l] *= self._counts[l] / (self._counts[1] - 1)
         l, w = scipy.linalg.eig(self._covs[0], self._covs[0] + self._covs[1])
+        l = l.real
         ind = l > 0
         l, w = l[ind], w[:, ind]
         ord = np.abs(l - 0.5).argsort()[::-1]

src/eegdash/features/feature_bank/connectivity.py

@@ -2,7 +2,7 @@
 import numba as nb
 from scipy import special
 
-from ..decorators import FeaturePredecessor, univariate_feature
+from ..decorators import univariate_feature
 from .signal import signal_zero_crossings
 
 
@@ -15,7 +15,6 @@
 ]
 
 
-@FeaturePredecessor()
 @univariate_feature
 @nb.njit(cache=True, fastmath=True)
 def dimensionality_higuchi_fractal_dim(x, k_max=10, eps=1e-7):
@@ -34,15 +33,13 @@ def dimensionality_higuchi_fractal_dim(x, k_max=10, eps=1e-7):
     return hfd
 
 
-@FeaturePredecessor()
 @univariate_feature
 def dimensionality_petrosian_fractal_dim(x):
     nd = signal_zero_crossings(np.diff(x, axis=-1))
     log_n = np.log(x.shape[-1])
     return log_n / (np.log(nd) + log_n)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def dimensionality_katz_fractal_dim(x):
     dists = np.abs(np.diff(x, axis=-1))
@@ -53,7 +50,6 @@ def dimensionality_katz_fractal_dim(x):
     return log_n / (np.log(d / L) + log_n)
 
 
-@FeaturePredecessor()
 @univariate_feature
 @nb.njit(cache=True, fastmath=True)
 def _hurst_exp(x, ns, a, gamma_ratios, log_n):
@@ -80,7 +76,6 @@ def _hurst_exp(x, ns, a, gamma_ratios, log_n):
     return h
 
 
-@FeaturePredecessor()
 @univariate_feature
 def dimensionality_hurst_exp(x):
     ns = np.unique(np.power(2, np.arange(2, np.log2(x.shape[-1]) - 1)).astype(int))
@@ -94,7 +89,6 @@ def dimensionality_hurst_exp(x):
     return _hurst_exp(x, ns, a, gamma_ratios, log_n)
 
 
-@FeaturePredecessor()
 @univariate_feature
 @nb.njit(cache=True, fastmath=True)
 def dimensionality_detrended_fluctuation_analysis(x):

src/eegdash/features/feature_bank/csp.py

@@ -2,12 +2,10 @@
 import numpy as np
 from scipy import stats
 
-from ..extractors import FeatureExtractor
-from ..decorators import FeaturePredecessor, univariate_feature
+from ..decorators import univariate_feature
 
 
 __all__ = [
-    "HjorthFeatureExtractor",
     "signal_mean",
     "signal_variance",
     "signal_skewness",
@@ -25,61 +23,51 @@
 ]
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_mean(x):
     return x.mean(axis=-1)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_variance(x, **kwargs):
     return x.var(axis=-1, **kwargs)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_std(x, **kwargs):
     return x.std(axis=-1, **kwargs)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_skewness(x, **kwargs):
     return stats.skew(x, axis=x.ndim - 1, **kwargs)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_kurtosis(x, **kwargs):
     return stats.kurtosis(x, axis=x.ndim - 1, **kwargs)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_root_mean_square(x):
     return np.sqrt(np.power(x, 2).mean(axis=-1))
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_peak_to_peak(x, **kwargs):
     return np.ptp(x, axis=-1, **kwargs)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_quantile(x, q: numbers.Number = 0.5, **kwargs):
     return np.quantile(x, q=q, axis=-1, **kwargs)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_line_length(x):
     return np.abs(np.diff(x, axis=-1)).mean(axis=-1)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_zero_crossings(x, threshold=1e-15):
     zero_ind = np.logical_and(x > -threshold, x < threshold)
@@ -90,25 +78,16 @@ def signal_zero_crossings(x, threshold=1e-15):
     return zero_cross
 
 
-@FeaturePredecessor()
-class HjorthFeatureExtractor(FeatureExtractor):
-    def preprocess(self, x):
-        return (x, np.diff(x, axis=-1), x.std(axis=-1))
-
-
-@FeaturePredecessor(HjorthFeatureExtractor)
 @univariate_feature
-def signal_hjorth_mobility(x, dx, x_std):
-    return dx.std(axis=-1) / x_std
+def signal_hjorth_mobility(x):
+    return np.diff(x, axis=-1).std(axis=-1) / x.std(axis=-1)
 
 
-@FeaturePredecessor(HjorthFeatureExtractor)
 @univariate_feature
-def signal_hjorth_complexity(x, dx, x_std):
-    return np.diff(dx, axis=-1).std(axis=-1) / x_std
+def signal_hjorth_complexity(x):
+    return np.diff(x, 2, axis=-1).std(axis=-1) / x.std(axis=-1)
 
 
-@FeaturePredecessor()
 @univariate_feature
 def signal_decorrelation_time(x, fs=1):
     f = np.fft.fft(x - x.mean(axis=-1, keepdims=True), axis=-1)

src/eegdash/features/feature_bank/dimensionality.py

@@ -22,7 +22,6 @@
 ]
 
 
-@FeaturePredecessor()
 class SpectralFeatureExtractor(FeatureExtractor):
     def preprocess(self, x, **kwargs):
         f_min = kwargs.pop("f_min") if "f_min" in kwargs else None