Skip to content

Latest commit

 

History

History
131 lines (62 loc) · 4.01 KB

operator_nodes.md

File metadata and controls

131 lines (62 loc) · 4.01 KB
Raw

Operator Nodes

Operator nodes implement functionson some input values. The following provides an overview of the different types available and links to the full documentation in the operator_nodes subdirectory. Arguments to an operator node's __init__ function may be used as keys in the dictionaries defining them in the simulation configuration file.

Conditional Operators

Implement conditional functions.

Classes:

  • IfElse

Constant Operators

Generate values representing constants and broadcast to some shape.

Classes:

  • Constant: A node that generates constant values.

  • RandomConstant: A node that generates constant values, where values are initially drawn from a distribution.

Distribution Operators

Function nodes that draw values from a distribution.

Classes:

  • Distribution: Draw values from a distribution.

Haplotype Combine Operators

Combine haplotypes into a single array.

Classes:

  • AdditiveCombine: A node that sums the two haplotypes element-wise to simulate an additive inheritance effect.

  • MaxCombine: A node that takes the element-wise maximum of the two haplotypes to simulate a dominant inheritance effect.

  • MinCombine: A node that takes the element-wise minimum of the two haplotypes to simulate a recessive inheritance effect.

  • MeanCombine: A node that takes the element-wise mean of the two haplotypes. Mean is either arithmetic, geometric, or harmonic.

Math Operators

General math function nodes for the simulation.

Classes:

  • Identity: A node that returns the input. Mainly for testing.

  • Sum: A node that sums some inputs element-wise.

  • Product: A node that multiplies some inputs element-wise.

Noise Operators

Operators that add noise in the phenotype simulation.

Classes:

  • GaussianNoise: A node that adds Gaussian noise to the input.

  • Heritability: Contols what fraction of the information output by the node is a function of the input vs. Gaussian noise. The output is a weighted average of the input and Gaussian noise. The Gaussian noise is scaled so that the output has the same variance as the input.

Reduce Operators

Reduce a Values matrix to one value per sample using some operation.

Classes:

  • SumReduce: Sum each sample's feature values into a single value per sample.

  • MinReduce: Return the minimum of each sample's feature values as the sample's value.

  • MaxReduce: Return the maximum of each sample's feature values as the sample's value.

  • MeanReduce: Return the mean of each sample's feature values as the sample's value. Mean is either arithmetic (default), geometric, or harmonic.

  • AnyReduce: Return 1 if any feature value is past a threshold, 0 otherwise.

  • AllReduce: Return 1 if all feature values are past a threshold, 0 otherwise.

Scaling Operators

Operators that scaling input distributions in the phenotype simulation.

Classes:

  • Clip: A node that clips the input to be greater than or equal to some minimum value and/or less than or equal to some maximum value.

  • MinMaxScaler: A node that scales the input to be between 0 and 1 using the minimum and maximum values of the input.

  • StandardScaler: A node that scales the input to have mean 0 and standard deviation 1 using the mean and standard deviation of the input.

  • RobustScaler: A node that scales the input to have median 0 and interquartile range 1 using the median and interquartile range of the input.

Transformation Function Operators

Non-linear transformation function operator nodes.

Take some input values and apply some transformation function, like a sigmoid, tanh, ReLU,

Classes:

  • ReLU

  • Sigmoid

  • Softmax

  • Tanh

Utility Operators

Utility functions.

Classes:

  • Concatenate: A node that concatenates some inputs into a single array.