compute_roc_curve_features: Computes the ROC curve by assessing the overlap of selected...

In lcrawlab/SINATRA: Sub Image Analysis using Topological Summary Statistics (SINATRA)

Computes the ROC curve by assessing the overlap of selected features and causal features.

Description

We compute the ROC curve by assessing the overlap of selected features and causal features. A causal feature here means to be associated to more than the (min_points) number of causal vertices. We do this for every complex in the data set then average the ROC curves.

Usage

compute_roc_curve_features(
  data,
  class_1_causal_points,
  class_2_causal_points,
  distance_to_causal_point = 0.1,
  rate_values,
  grid_size,
  eta = 0.1,
  directions_per_cone,
  curve_length,
  truncated = -1,
  class = 0,
  ball = TRUE,
  ball_radius = ball_radius,
  dir,
  min_points = 2,
  mode = "grid",
  subdivision = 3
)

Arguments

data	(list) : Metadata about the simulated shapes (vertex coordinates, etc.)
class_1_causal_points	: Vertex indices of causal points for class 1.
class_2_causal_points	: Vertex indices of causal points for class 2.
distance_to_causal_point	(float) : For interpolated shapes, the distance from a vertex to the causal points to be considered a "causal vertex"
rate_values	(vector) : Vector of variable importances for each sub-level set across each direction in a given cone.
grid_size	(int) : The fine-ness/granularity of the interpolated shapes.
eta	(float) : The kernel shape parameter.
directions_per_cone	(int): The number of directions we want generated within each cone.
curve_length	(int) : Number of sub-level sets in each EC computation.
truncated	(int) : The number of "cuts" to compute TPR/FPR for the ROC curve over. Used to speed up ROC computations.
class	(int) : The class of the group of shapes we compute the ROC curve against.
ball	(boolean) : Denotes whether or not to compute the EC curves over a ball for uniform measurements
ball_radius	(float) : The radius of the bounding ball used if we compute the balled EC curve.
dir	(nx3 matrix): The matrix of directions for which the (S/D) EC curve were computed over.
min_points	(int) : Used when type = 'feature'. The mininum number of causal vertices for a sub-level set to be associated with to be considered a causal 'feature'.
mode	(string) : The data generation scheme. We currently support 'sphere', 'gaussian_grid", or interpolations (default).
subdivision	(int) : The fineness of the sphere meshes (if mode == 'sphere'). We currently use subdivision = 3.

Value

total_rate_roc (matrix) : The ROC curve.

lcrawlab/SINATRA documentation built on Sept. 13, 2023, 2 p.m.