Conditional density estimation is a longstanding and challenging problem in statistical theory, and numerous proposals exist for optimally estimating such complex functions. An algorithm for nonparametric estimation of conditional densities based on a pooled hazard regression formulation is implemented based on the highly adaptive lasso (HAL), an optimal algorithm for efficient estimation with guarantees of fast convergence rates.

haldensify()

References

Documentation:

https://code.nimahejazi.org/haldensify

GitHub repository:

https://github.com/nhejazi/haldensify