This repository contains an implementation of the algorithm presented in
*Black-box density function estimation using recursive partitioning*
[arXiv](https://arxiv.org/abs/2010.13632), referred to as DEFER (DEnsity Function Estimation using Recursive partitioning). The paper was presented at the International Conference on Machine Learning (ICML) 2021.
DEFER allows efficient Bayesian inference on general problems involving up to about ten random variables or dimensions,
without the need to specify anything else than
- the unnormalised density function (which can be black-box),
- domain bounds (enclosing the typical set, although allowed to be doing so with very large margins),
- a density function evaluation budget.
The code comes with a high-level interface intended for ease-of-use.
Below we will show:
- How the DEFER algorithm can be used to approximate arbitrary distributions (of moderate dimension),
such as posterior distributions which are allowed to be e.g. multimodal, discontinuous, exhibit complicated correlations, and have zero density regions.
- Some available operations the approximation provides, including:
- fast, constant time sampling,
- computation of normalisation constant,
- analytical expectations of functions with respect to the approximation,
- conditionals to be derived quickly, and
- approximate marginalisation of a subset of variables or dimensions.
```python
import corner
import matplotlib.pylab as plt
from defer.helpers import *
from defer.variables import Variable
import matplotlib.pyplot as plt
import warnings
warnings.filterwarnings('ignore')
# Configure plot defaults
plt.rcParams['axes.facecolor'] = 'white'
plt.rcParams['grid.color'] = '#666666'
%config InlineBackend.figure_format = 'png'
```
### Density function example
The DEFER algorithm constructs a distribution approximation given a provided (arbitrary) unnormalised density function $f$ which can be point-wise evaluated.
In this example we provide the algorithm with such a function, namely where the density value at a given location depends on the distance to the closest point of a shape embedded in the same space. The particular shape used is a spiral embedded in three dimensions. The distance (as a function of $x$) is used as the energy function of a Gibbs distribution.
Note that the DEFER algorithm does not need to know what the function $f$ is or represents; it will simply query it with inputs and construct an approximation which allows operations that typically is not available using the original function $f$ (except for trivial $f$ and corresponding distributions).
```python
def spiral(t):
w = 6 * np.pi
v = 0.9
c = 1
return np.array([
t,
(v * t + c) * np.cos(w * t),
(v * t + c) * np.sin(w * t),
]).transpose()
# Domain bounds
curve_lower_x = -1
curve_upper_x = 1
other_dims_lower = [-2, -2]
other_dims_upper = [2, 2]
lower = [curve_lower_x] + other_dims_lower
upper = [curve_upper_x] + other_dims_upper
def find_shortest_distance(
x, curve, lower_x=curve_lower_x, upper_x=curve_upper_x):
trial_input_points = np.linspace(lower_x, upper_x, num=5000)
curve_points = curve(trial_input_points)
distances = np.linalg.norm(np.abs(curve_points - x), axis=1)
index = int(np.argmin(distances))
shortest_distance = distances[index]
return shortest_distance
# Unnormalised density function. This can be any continuous-input function returning a density value.
def f(x):
distance_to_closest_point_on_curve = find_shortest_distance(x, curve=spiral)
energy = distance_to_closest_point_on_curve
return np.exp(-30 * energy)
```
#### Variables
Declare the random variables, corresponding to the parameters of the density function $f$.
In this example, we only have a single parameter $x$, and thus only one random variable.
The Variables object produces the density function domain (variables.domain) - or sample space. It will also keep track of mappings between variables (or slices of them using []) to the corresponding dimensions.
Note that the variables do not have any state - they simply act as semantic collections of dimensions with defined boundaries. Having the variables kept together as a collection in the Variables object will be particularly useful for keeping track of what remains when deriving conditionals or marginals (more on that later).
```python
x = Variable(
lower=lower,
upper=upper,
name="x"
)
variables = Variables([x])
```
#### Construct approximation
Here we use a high-level wrapper around the DEFER algorithm which given the $f$-function definition and the variables (setting the domain boundaries), returns a distribution approximation.
The budget of the algorithm, in terms of number of evaluations of $f$, is set using 'num_fn_calls'. Note that this number may differ slightly from the calls actually used (by a few evaluations). This is because the construction of each partition (see paper) requires more than one density evaluations, and thus the total may not add up exactly to the provided number.
Note that 'is_log_fn' also needs to be specified. 'is_log_fn' should be set to True if the $f$ function returns the logarithm of the density value rather than the density value.
```python
approx_joint: DensityFunctionApproximation = construct(
fn=f,
variables=variables,
is_log_fn=False,
num_fn_calls=30000,
callback=lambda i, density:
print("#Evals: %s. Log Z: %.2f" %
(density.num_partitions, np.log(density.z))),
callback_freq_fn_calls=3000,
)
```
#Evals: 7. Log Z: -4.71
#Evals: 3015. Log Z: -1.58
#Evals: 6019. Log Z: -1.48
#Evals: 9025. Log Z: -1.46
#Evals: 12027. Log Z: -1.41
#Evals: 15027. Log Z: -1.41
#Evals: 18039. Log Z: -1.41
#Evals: 21043. Log Z: -1.41
#Evals: 24063. Log Z: -1.40
#Evals: 27063. Log Z: -1.39
#Evals: 30001. Log Z: -1.39
#### The DensityFunctionApproximation class
The algorithm returns a 'DensityFunctionApproximation' object. This is a high-level wrapper around the approximation, which is a piece-wise constant function with as many elements/partitions as number of $f$ evaluations made. The object provides several often useful methods and properties which we will look at now.
Normalisation constant of the approximation. This can be used for, e.g., evidence estimation.
```python
z = approx_joint.z
```
Analytical expectation of functions with respect to the distribution approximation.
Due to the piece-wise constant approximation being used, this simply corresponds to a weighted summation. The 'expectation' method is for convenience for computing expectations of functions provided the function, taking the integrand ($f$) value, the corresponding parameters (concatenated to a vector), and the normalisation constant $z$, as parameters. In the below example the differential entropy is computed this way.
```python
differential_entropy = approx_joint.expectation(lambda f, x, z: -np.log(f / z))
```
A few, very common such expectations are provided as methods (such as mean and variance), which are internally using the 'expectation' method.
```python
mean = approx_joint.mean()
var = approx_joint.var()
```
The (estimated) mode is provided by the 'mode' method. This is simply the centroid of the partition of the approximation which has the highest associated $f$ value.
```python
mode = approx_joint.mode()
```
The approximation of the density function can be queried just as the original function $f$ by calling with the same parameter(s). A normalised version (the probability density) is provided using the 'prob' method.
```python
x_test = np.array([0.5, 0.5, 0.5])
f_test = approx_joint(x_test)
p_test = approx_joint.prob(x_test)
```
Often what one is interested in is posterior samples.
To obtain samples from the distribution approximation, we first construct a sampler function using the 'sampler' method. Internally it performs the pre-processing required to be able to sample very quickly.
```python
sampler = approx_joint.sampler()
```
The sampler function allows sampling from the distribution in constant time per sample.
Below we draw one million samples.
Returned is a list of numpy arrays - one array of samples per variable.
In our example, however, we only have the $x$ variable (with three dimensions).
```python
x_samples, = sampler(num_samples=1000000)
x_samples.shape
```
(1000000, 3)
The DensityFunctionApproximation can easily be saved or loaded using the 'save' and 'load' methods. Internally the tree-structure of partitions (only) are saved/loaded, instead of the whole Python object.
```python
approx_joint.save("/tmp/approx_joint.pickle")
approx_joint.load("/tmp/approx_joint.pickle")
```
We will now plot the samples in a corner plot in order to visualize the distribution.
```python
def plot(density: DensityFunctionApproximation, samples=None):
if samples is None:
print("Preparing sampler..")
sampler = density.sampler()
print("Sampling..")
samples_per_variable = sampler(num_samples=10 ** 6)
samples = np.concatenate(samples_per_variable, axis=-1)
print("Plotting..")
figure = corner.corner(
samples,
range=density.variables.bounds,
labels=[
"%s: dim %s" % (var.name, index)
for var in density.variables.variable_slices
for index in var.indices
],
plot_contours=False,
no_fill_contours=True,
bins=100,
plot_datapoints=False,
)
plt.show()
plot(approx_joint, samples=x_samples)
```
Plotting..
#### Conditionals
We can easily derive a conditional using the 'conditional' method,
yielding a new approximation of the same type.
In this example we condition the last dimension of the $x$ variable to 0.5.
```python
approx_conditional: DensityFunctionApproximation = approx_joint.conditional({
x[-1]: np.array([0.5])
})
```
Let us plot samples from the derived conditional in a corner plot to visualize it.
```python
plot(approx_conditional)
```
Preparing sampler..
Sampling..
Plotting..
The conditional of the approximation, being of the same type, provides the same methods and properties.
Including its normalisation constant, mode, and methods for returning density values and computing expectations.
```python
z = approx_conditional.z
approx_conditional.mode()
approx_conditional.expectation(lambda f, x, z: -np.log(f / z))
approx_conditional.mean()
approx_conditional.var()
approx_conditional.prob(np.array([0.5, 0.5]))
```
7.837536348532179338e-08
#### Marginalisation
We may also marginalise variables (or dimensions of variables) using the 'construct_marginal' helper function. Internally it treats marginalisation as just another problem for the DEFER algorithm: now where each $f$ evaluation requires (approximate) integration over some variables - which using this helper function is performed using DEFER as well, i.e. in an inner-loop.
Returned is a DensityFunctionApproximation similar to in the previously discussed cases.
Note that in this example the provided $f$ is the constructed approximation of $f$, rather than $f$ itself. This can allow a significant speed-up if $f$ takes relatively much time to evaluate (more than several milliseconds). Alternatively $f$ can be passed in instead.
In the below example the first dimension of the $x$ variable is marginalised, returning a distribution now only two dimensions rather than three.
```python
approx_marginal: DensityFunctionApproximation = construct_marginal(
fn=approx_joint,
variables=variables,
marginalize_variable_slices=[x[0]],
is_log_fn=False,
num_outer_fn_calls=5000,
num_inner_fn_calls=15,
callback=lambda i, density:
print("#Evals: %s. Log Z: %.2f" %
(density.num_partitions, np.log(density.z))),
callback_freq_fn_calls=500,
)
```
#Evals: 5. Log Z: -1.16
#Evals: 509. Log Z: -1.48
#Evals: 1009. Log Z: -1.48
#Evals: 1509. Log Z: -1.46
#Evals: 2011. Log Z: -1.45
#Evals: 2515. Log Z: -1.44
#Evals: 3019. Log Z: -1.44
#Evals: 3523. Log Z: -1.44
#Evals: 4027. Log Z: -1.43
#Evals: 4531. Log Z: -1.43
#Evals: 5007. Log Z: -1.43
```python
plot(approx_marginal)
```
Preparing sampler..
Sampling..
Plotting..
```python
approx_marginal.z
approx_marginal.mode()
approx_marginal.expectation(lambda f, x, z: -np.log(f / z))
approx_marginal.mean()
approx_marginal.var()
approx_marginal.prob(np.array([0.5, 0.5]))
```
0.060272058916242459574
#### Combining operations arbitrarily.
As the marginal approximation provides the same operations,
we can of course, for example, derive a conditional of the marginal similar to before - or use any of the other operations, in any order.
```python
approx_conditional_of_marginal = approx_marginal.conditional({
x[-1]: np.array([0.5])
})
```
```python
plot(approx_conditional_of_marginal)
```
Preparing sampler..
Sampling..
Plotting..
```python
approx_conditional_of_marginal.z
approx_conditional_of_marginal.mode()
approx_conditional_of_marginal.expectation(lambda f, x, z: -np.log(f / z))
approx_conditional_of_marginal.mean()
approx_conditional_of_marginal.var()
approx_conditional_of_marginal.prob(np.array([0.5]))
```
0.18142085889723261718
Raw data
{
"_id": null,
"home_page": null,
"name": "deference",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": null,
"keywords": "density estimation, bayesian inference, machine learning, statistics",
"author": null,
"author_email": "Erik Bodin <mail@erikbodin.com>",
"download_url": "https://files.pythonhosted.org/packages/0f/89/bde4b67986d2b30703af85ed561b476e63b1b1ef089857e9b78658e6d27a/deference-0.1.1.tar.gz",
"platform": null,
"description": "This repository contains an implementation of the algorithm presented in\n*Black-box density function estimation using recursive partitioning*\n[arXiv](https://arxiv.org/abs/2010.13632), referred to as DEFER (DEnsity Function Estimation using Recursive partitioning). The paper was presented at the International Conference on Machine Learning (ICML) 2021.\n\nDEFER allows efficient Bayesian inference on general problems involving up to about ten random variables or dimensions,\nwithout the need to specify anything else than\n - the unnormalised density function (which can be black-box),\n - domain bounds (enclosing the typical set, although allowed to be doing so with very large margins),\n - a density function evaluation budget.\n\nThe code comes with a high-level interface intended for ease-of-use.\n\nBelow we will show:\n- How the DEFER algorithm can be used to approximate arbitrary distributions (of moderate dimension),\nsuch as posterior distributions which are allowed to be e.g. multimodal, discontinuous, exhibit complicated correlations, and have zero density regions.\n- Some available operations the approximation provides, including:\n - fast, constant time sampling,\n - computation of normalisation constant,\n - analytical expectations of functions with respect to the approximation,\n - conditionals to be derived quickly, and\n - approximate marginalisation of a subset of variables or dimensions.\n\n\n```python\nimport corner\nimport matplotlib.pylab as plt\nfrom defer.helpers import *\nfrom defer.variables import Variable\nimport matplotlib.pyplot as plt\nimport warnings\nwarnings.filterwarnings('ignore')\n# Configure plot defaults\nplt.rcParams['axes.facecolor'] = 'white'\nplt.rcParams['grid.color'] = '#666666'\n%config InlineBackend.figure_format = 'png'\n```\n\n### Density function example\nThe DEFER algorithm constructs a distribution approximation given a provided (arbitrary) unnormalised density function $f$ which can be point-wise evaluated. \n\nIn this example we provide the algorithm with such a function, namely where the density value at a given location depends on the distance to the closest point of a shape embedded in the same space. The particular shape used is a spiral embedded in three dimensions. The distance (as a function of $x$) is used as the energy function of a Gibbs distribution.\n\nNote that the DEFER algorithm does not need to know what the function $f$ is or represents; it will simply query it with inputs and construct an approximation which allows operations that typically is not available using the original function $f$ (except for trivial $f$ and corresponding distributions).\n\n\n```python\ndef spiral(t):\n w = 6 * np.pi\n v = 0.9\n c = 1\n return np.array([\n t,\n (v * t + c) * np.cos(w * t),\n (v * t + c) * np.sin(w * t),\n ]).transpose()\n\n# Domain bounds\ncurve_lower_x = -1\ncurve_upper_x = 1\nother_dims_lower = [-2, -2]\nother_dims_upper = [2, 2]\nlower = [curve_lower_x] + other_dims_lower\nupper = [curve_upper_x] + other_dims_upper\n\ndef find_shortest_distance(\n x, curve, lower_x=curve_lower_x, upper_x=curve_upper_x):\n trial_input_points = np.linspace(lower_x, upper_x, num=5000)\n curve_points = curve(trial_input_points)\n distances = np.linalg.norm(np.abs(curve_points - x), axis=1)\n index = int(np.argmin(distances))\n shortest_distance = distances[index]\n return shortest_distance\n\n# Unnormalised density function. This can be any continuous-input function returning a density value.\ndef f(x):\n distance_to_closest_point_on_curve = find_shortest_distance(x, curve=spiral)\n energy = distance_to_closest_point_on_curve\n return np.exp(-30 * energy)\n```\n\n#### Variables\nDeclare the random variables, corresponding to the parameters of the density function $f$.\nIn this example, we only have a single parameter $x$, and thus only one random variable.\n\nThe Variables object produces the density function domain (variables.domain) - or sample space. It will also keep track of mappings between variables (or slices of them using []) to the corresponding dimensions.\n\nNote that the variables do not have any state - they simply act as semantic collections of dimensions with defined boundaries. Having the variables kept together as a collection in the Variables object will be particularly useful for keeping track of what remains when deriving conditionals or marginals (more on that later).\n\n\n```python\nx = Variable(\n lower=lower,\n upper=upper,\n name=\"x\"\n)\nvariables = Variables([x])\n```\n\n#### Construct approximation\nHere we use a high-level wrapper around the DEFER algorithm which given the $f$-function definition and the variables (setting the domain boundaries), returns a distribution approximation.\n\nThe budget of the algorithm, in terms of number of evaluations of $f$, is set using 'num_fn_calls'. Note that this number may differ slightly from the calls actually used (by a few evaluations). This is because the construction of each partition (see paper) requires more than one density evaluations, and thus the total may not add up exactly to the provided number.\n\nNote that 'is_log_fn' also needs to be specified. 'is_log_fn' should be set to True if the $f$ function returns the logarithm of the density value rather than the density value.\n\n\n```python\napprox_joint: DensityFunctionApproximation = construct(\n fn=f,\n variables=variables,\n is_log_fn=False,\n num_fn_calls=30000,\n callback=lambda i, density:\n print(\"#Evals: %s. Log Z: %.2f\" %\n (density.num_partitions, np.log(density.z))),\n callback_freq_fn_calls=3000,\n)\n```\n\n #Evals: 7. Log Z: -4.71\n #Evals: 3015. Log Z: -1.58\n #Evals: 6019. Log Z: -1.48\n #Evals: 9025. Log Z: -1.46\n #Evals: 12027. Log Z: -1.41\n #Evals: 15027. Log Z: -1.41\n #Evals: 18039. Log Z: -1.41\n #Evals: 21043. Log Z: -1.41\n #Evals: 24063. Log Z: -1.40\n #Evals: 27063. Log Z: -1.39\n #Evals: 30001. Log Z: -1.39\n\n\n#### The DensityFunctionApproximation class\nThe algorithm returns a 'DensityFunctionApproximation' object. This is a high-level wrapper around the approximation, which is a piece-wise constant function with as many elements/partitions as number of $f$ evaluations made. The object provides several often useful methods and properties which we will look at now.\n\nNormalisation constant of the approximation. This can be used for, e.g., evidence estimation.\n\n\n```python\nz = approx_joint.z\n```\n\nAnalytical expectation of functions with respect to the distribution approximation.\nDue to the piece-wise constant approximation being used, this simply corresponds to a weighted summation. The 'expectation' method is for convenience for computing expectations of functions provided the function, taking the integrand ($f$) value, the corresponding parameters (concatenated to a vector), and the normalisation constant $z$, as parameters. In the below example the differential entropy is computed this way.\n\n\n```python\ndifferential_entropy = approx_joint.expectation(lambda f, x, z: -np.log(f / z))\n```\n\nA few, very common such expectations are provided as methods (such as mean and variance), which are internally using the 'expectation' method.\n\n\n```python\nmean = approx_joint.mean()\nvar = approx_joint.var()\n```\n\nThe (estimated) mode is provided by the 'mode' method. This is simply the centroid of the partition of the approximation which has the highest associated $f$ value.\n\n\n```python\nmode = approx_joint.mode()\n```\n\nThe approximation of the density function can be queried just as the original function $f$ by calling with the same parameter(s). A normalised version (the probability density) is provided using the 'prob' method.\n\n\n```python\nx_test = np.array([0.5, 0.5, 0.5])\nf_test = approx_joint(x_test)\np_test = approx_joint.prob(x_test)\n```\n\nOften what one is interested in is posterior samples. \nTo obtain samples from the distribution approximation, we first construct a sampler function using the 'sampler' method. Internally it performs the pre-processing required to be able to sample very quickly.\n\n\n```python\nsampler = approx_joint.sampler()\n```\n\nThe sampler function allows sampling from the distribution in constant time per sample.\nBelow we draw one million samples. \nReturned is a list of numpy arrays - one array of samples per variable. \nIn our example, however, we only have the $x$ variable (with three dimensions).\n\n\n```python\nx_samples, = sampler(num_samples=1000000)\nx_samples.shape\n```\n\n\n\n\n (1000000, 3)\n\n\n\nThe DensityFunctionApproximation can easily be saved or loaded using the 'save' and 'load' methods. Internally the tree-structure of partitions (only) are saved/loaded, instead of the whole Python object.\n\n\n```python\napprox_joint.save(\"/tmp/approx_joint.pickle\")\napprox_joint.load(\"/tmp/approx_joint.pickle\")\n```\n\nWe will now plot the samples in a corner plot in order to visualize the distribution.\n\n\n```python\ndef plot(density: DensityFunctionApproximation, samples=None):\n if samples is None:\n print(\"Preparing sampler..\")\n sampler = density.sampler()\n print(\"Sampling..\")\n samples_per_variable = sampler(num_samples=10 ** 6)\n samples = np.concatenate(samples_per_variable, axis=-1)\n print(\"Plotting..\")\n figure = corner.corner(\n samples,\n range=density.variables.bounds,\n labels=[\n \"%s: dim %s\" % (var.name, index)\n for var in density.variables.variable_slices\n for index in var.indices\n ],\n plot_contours=False,\n no_fill_contours=True,\n bins=100,\n plot_datapoints=False,\n )\n plt.show()\n\nplot(approx_joint, samples=x_samples)\n```\n\n Plotting..\n\n\n\n\n\n\n#### Conditionals\nWe can easily derive a conditional using the 'conditional' method,\nyielding a new approximation of the same type.\nIn this example we condition the last dimension of the $x$ variable to 0.5.\n\n\n```python\napprox_conditional: DensityFunctionApproximation = approx_joint.conditional({\n x[-1]: np.array([0.5])\n})\n```\n\nLet us plot samples from the derived conditional in a corner plot to visualize it.\n\n\n```python\nplot(approx_conditional)\n```\n\n Preparing sampler..\n Sampling..\n Plotting..\n\n\n\n\n\n\nThe conditional of the approximation, being of the same type, provides the same methods and properties. \nIncluding its normalisation constant, mode, and methods for returning density values and computing expectations.\n\n\n```python\nz = approx_conditional.z\napprox_conditional.mode()\napprox_conditional.expectation(lambda f, x, z: -np.log(f / z))\napprox_conditional.mean()\napprox_conditional.var()\napprox_conditional.prob(np.array([0.5, 0.5]))\n```\n\n\n\n\n 7.837536348532179338e-08\n\n\n\n#### Marginalisation\nWe may also marginalise variables (or dimensions of variables) using the 'construct_marginal' helper function. Internally it treats marginalisation as just another problem for the DEFER algorithm: now where each $f$ evaluation requires (approximate) integration over some variables - which using this helper function is performed using DEFER as well, i.e. in an inner-loop. \n\nReturned is a DensityFunctionApproximation similar to in the previously discussed cases. \n\nNote that in this example the provided $f$ is the constructed approximation of $f$, rather than $f$ itself. This can allow a significant speed-up if $f$ takes relatively much time to evaluate (more than several milliseconds). Alternatively $f$ can be passed in instead.\n\nIn the below example the first dimension of the $x$ variable is marginalised, returning a distribution now only two dimensions rather than three.\n\n\n```python\napprox_marginal: DensityFunctionApproximation = construct_marginal(\n fn=approx_joint,\n variables=variables,\n marginalize_variable_slices=[x[0]],\n is_log_fn=False,\n num_outer_fn_calls=5000,\n num_inner_fn_calls=15,\n callback=lambda i, density:\n print(\"#Evals: %s. Log Z: %.2f\" %\n (density.num_partitions, np.log(density.z))),\n callback_freq_fn_calls=500,\n)\n```\n\n #Evals: 5. Log Z: -1.16\n #Evals: 509. Log Z: -1.48\n #Evals: 1009. Log Z: -1.48\n #Evals: 1509. Log Z: -1.46\n #Evals: 2011. Log Z: -1.45\n #Evals: 2515. Log Z: -1.44\n #Evals: 3019. Log Z: -1.44\n #Evals: 3523. Log Z: -1.44\n #Evals: 4027. Log Z: -1.43\n #Evals: 4531. Log Z: -1.43\n #Evals: 5007. Log Z: -1.43\n\n\n\n```python\nplot(approx_marginal)\n```\n\n Preparing sampler..\n Sampling..\n Plotting..\n\n\n\n\n\n\n\n```python\napprox_marginal.z\napprox_marginal.mode()\napprox_marginal.expectation(lambda f, x, z: -np.log(f / z))\napprox_marginal.mean()\napprox_marginal.var()\napprox_marginal.prob(np.array([0.5, 0.5]))\n```\n\n\n\n\n 0.060272058916242459574\n\n\n\n#### Combining operations arbitrarily.\nAs the marginal approximation provides the same operations, \nwe can of course, for example, derive a conditional of the marginal similar to before - or use any of the other operations, in any order.\n\n\n```python\napprox_conditional_of_marginal = approx_marginal.conditional({\n x[-1]: np.array([0.5])\n})\n```\n\n\n```python\nplot(approx_conditional_of_marginal)\n```\n\n Preparing sampler..\n Sampling..\n Plotting..\n\n\n\n\n\n\n\n```python\napprox_conditional_of_marginal.z\napprox_conditional_of_marginal.mode()\napprox_conditional_of_marginal.expectation(lambda f, x, z: -np.log(f / z))\napprox_conditional_of_marginal.mean()\napprox_conditional_of_marginal.var()\napprox_conditional_of_marginal.prob(np.array([0.5]))\n```\n\n\n\n\n 0.18142085889723261718\n\n\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "DEnsity Function Estimation using Recursive partitioning",
"version": "0.1.1",
"project_urls": {
"Documentation": "https://github.com/bodin-e/defer#readme",
"Homepage": "https://github.com/bodin-e/defer",
"Issues": "https://github.com/bodin-e/defer/issues",
"Repository": "https://github.com/bodin-e/defer"
},
"split_keywords": [
"density estimation",
" bayesian inference",
" machine learning",
" statistics"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "b61aaec7a069e047d6d2c00cf7939932a61540ca46cd3438ef017e75358114e3",
"md5": "39175d51737cae3e25ea73abba5d3978",
"sha256": "9fe2f0f73acac4444bb2b24bbbeeaacbd7cb449181b455388c900e2385789d94"
},
"downloads": -1,
"filename": "deference-0.1.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "39175d51737cae3e25ea73abba5d3978",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 32280,
"upload_time": "2025-09-04T16:14:38",
"upload_time_iso_8601": "2025-09-04T16:14:38.544381Z",
"url": "https://files.pythonhosted.org/packages/b6/1a/aec7a069e047d6d2c00cf7939932a61540ca46cd3438ef017e75358114e3/deference-0.1.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "0f89bde4b67986d2b30703af85ed561b476e63b1b1ef089857e9b78658e6d27a",
"md5": "632c8346bac4dda020e9324d9b05ef91",
"sha256": "6b2355b07789bff15f2e19908a0c1bcecbc5b59f84fff46b3a3ae7cd49847fe3"
},
"downloads": -1,
"filename": "deference-0.1.1.tar.gz",
"has_sig": false,
"md5_digest": "632c8346bac4dda020e9324d9b05ef91",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 770233,
"upload_time": "2025-09-04T16:14:40",
"upload_time_iso_8601": "2025-09-04T16:14:40.092278Z",
"url": "https://files.pythonhosted.org/packages/0f/89/bde4b67986d2b30703af85ed561b476e63b1b1ef089857e9b78658e6d27a/deference-0.1.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-09-04 16:14:40",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "bodin-e",
"github_project": "defer#readme",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [
{
"name": "numpy",
"specs": [
[
"==",
"1.19.5"
]
]
},
{
"name": "scipy",
"specs": [
[
"==",
"1.4.1"
]
]
},
{
"name": "scikit-spatial",
"specs": [
[
"==",
"5.0.0"
]
]
},
{
"name": "xxhash",
"specs": [
[
"==",
"1.4.3"
]
]
},
{
"name": "sobol-seq",
"specs": [
[
"==",
"0.1.2"
]
]
}
],
"lcname": "deference"
}
JSON
