harissa.inference.network module#

Core functions for network inference using likelihood maximization

Functions#

`grad_p1`	Smoothed L1 penalization gradient.
`grad_penalization`	Penalization gradient of network parameters.
`grad_theta`	Objective gradient (one time point).
`infer_network`	Network inference procedure.
`infer_proteins`	Estimate y directly from data.
`objective`	Objective function to be minimized (one time point).
`p1`	Smoothed L1 penalization.
`penalization`	Penalization of network parameters.

harissa.inference.network.grad_p1(x, s)[source]#: Smoothed L1 penalization gradient.

harissa.inference.network.grad_penalization(theta, theta0, t)[source]#: Penalization gradient of network parameters.

harissa.inference.network.grad_theta(theta, theta0, x, y, a, c, d, l, t)[source]#: Objective gradient (one time point).

harissa.inference.network.infer_network(x, y, a, c, l, tol, verb)[source]#: Network inference procedure.

harissa.inference.network.infer_proteins(x, a)[source]#: Estimate y directly from data.

harissa.inference.network.objective(theta, theta0, x, y, a, c, d, l, t)[source]#: Objective function to be minimized (one time point).

harissa.inference.network.p1(x, s)[source]#: Smoothed L1 penalization.

harissa.inference.network.penalization(theta, theta0, t)[source]#: Penalization of network parameters.