I am an archaeologist studying an ancient quarry comprised of a number of blocks. I have measurements of the size of these blocks. Generally these measurements should follow a LogNormal or Weibull distribution (i.e. positive, bigger blocks rarer etc.). I have reason to believe a portion of the obser…

To start with, I can tell you what is wrong with your code. In scipy when you do r = np.linspace(0,5,1000) y = scipy.stats.lognorm.pdf(r, 1,0) * (1 - 0.5*(scipy.stats.norm.pdf(r,2, 0.2))) This combines the probability distributions of the two random variables evaluated at the same observed value (…

So, I took a shot at using the model I suggested above: import numpy as np import scipy import matplotlib.pyplot as plt import pymc as pm import scipy.stats import arviz as az from pytensor.tensor import TensorVariable def sample(pdf, x, rng, N): cdf = np.cumsum(pdf) cdf = cdf / cdf[-1] val…

Thankyou very much for the detailed and helpful responses! Yes, I have considered identifiability to be a potential problem and I would expect the model to struggle if there is little selection going on. I am fine with restricting the priors a bit more such that it becomes less of an issue, I have …

Does this give some ideas? Modeling with soft truncation Sounds similar on the surface except with interval truncation instead of edge truncation Since you care about those bounds I would try to model them directly.

Modelling a distribution with a portion removed from it

Questions v5

kpjmcg February 23, 2024, 1:03am 6

Yes it does thanks. I’ll have a go at trying to produce two soft truncations.

Is it possible to model systems with random variables of unknown distribution?

Observed data with uncertainty described by an irregular distribution

Topic		Replies	Views
Applying a selection function Questions	5	460	September 18, 2022
Truncated normal distribution with optimal truncation thresholds via DensityDist Questions	8	3113	June 21, 2018
Modeling with soft truncation version agnostic modeling	6	511	December 24, 2023
Censoring a truncated distribution v5	7	484	November 22, 2023
Is this a censored-data problem or not? version agnostic	6	312	October 27, 2023

Modelling a distribution with a portion removed from it

Related topics