Source code for pymc.tuning.scaling

#   Copyright 2024 The PyMC Developers
#
#   Licensed under the Apache License, Version 2.0 (the "License");
#   you may not use this file except in compliance with the License.
#   You may obtain a copy of the License at
#
#       http://www.apache.org/licenses/LICENSE-2.0
#
#   Unless required by applicable law or agreed to in writing, software
#   distributed under the License is distributed on an "AS IS" BASIS,
#   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
#   See the License for the specific language governing permissions and
#   limitations under the License.

import numpy as np

from numpy import exp, log, sqrt

from pymc.blocking import DictToArrayBijection
from pymc.model import Point, modelcontext
from pymc.pytensorf import hessian_diag
from pymc.util import get_var_name

__all__ = ["find_hessian", "trace_cov", "guess_scaling"]


def fixed_hessian(point, model=None):
    """
    Returns a fixed Hessian for any chain location.

    Parameters
    ----------
    model: Model (optional if in `with` context)
    point: dict
    vars: list
        Variables for which Hessian is to be calculated.
    """

    model = modelcontext(model)
    point = Point(point, model=model)

    rval = np.ones(DictToArrayBijection.map(point).size) / 10
    return rval


[docs] def find_hessian(point, vars=None, model=None, negate_output=True): """ Returns Hessian of logp at the point passed. Parameters ---------- model: Model (optional if in `with` context) point: dict vars: list Variables for which Hessian is to be calculated. """ model = modelcontext(model) H = model.compile_d2logp(vars, negate_output=negate_output) return H(Point(point, filter_model_vars=True, model=model))
def find_hessian_diag(point, vars=None, model=None, negate_output=True): """ Returns Hessian of logp at the point passed. Parameters ---------- model: Model (optional if in `with` context) point: dict vars: list Variables for which Hessian is to be calculated. """ model = modelcontext(model) H = model.compile_fn(hessian_diag(model.logp(), vars, negate_output=negate_output)) return H(Point(point, model=model)) def guess_scaling(point, vars=None, model=None, scaling_bound=1e-8): model = modelcontext(model) try: h = -find_hessian_diag(point, vars, model=model, negate_output=False) except NotImplementedError: h = fixed_hessian(point, model=model) return adjust_scaling(h, scaling_bound) def adjust_scaling(s, scaling_bound): if s.ndim < 2: return adjust_precision(s, scaling_bound) else: val, vec = np.linalg.eigh(s) val = adjust_precision(val, scaling_bound) return eig_recompose(val, vec) def adjust_precision(tau, scaling_bound=1e-8): mag = sqrt(abs(tau)) bounded = bound(log(mag), log(scaling_bound), log(1.0 / scaling_bound)) return exp(bounded) ** 2 def bound(a, l, u): # noqa E741 return np.maximum(np.minimum(a, u), l) def eig_recompose(val, vec): return vec.dot(np.diag(val)).dot(vec.T) def trace_cov(trace, vars=None, model=None): """ Calculate the flattened covariance matrix using a sample trace Useful if you want to base your covariance matrix for further sampling on some initial samples. Parameters ---------- trace: Trace vars: list variables for which to calculate covariance matrix Returns ------- r: array (n,n) covariance matrix """ model = modelcontext(model) if model is not None: vars = model.free_RVs elif vars is None: vars = trace.varnames def flat_t(var): x = trace[get_var_name(var)] return x.reshape((x.shape[0], np.prod(x.shape[1:], dtype=int))) return np.cov(np.concatenate(list(map(flat_t, vars)), 1).T)