Comments (3)
I'm not sure what you mean by "I have ensured that these small differences are not MCMC randomness", but this seems totally consistent to me! ArviZ estimates the MC error on the means as 0.001, 0.006, and 0.003, respectively so these differences are well within what you would expect for Monte Carlo sampling error.
More formally: the factor of -0.5*n*log(2*pi)
is an additive constant to the likelihood and this cannot have an effect on the results since MCMC algorithms only ever evaluate differences of log probabilities. There can, however, be numerical differences introduced by floating point precision that can affect the specific traces when adding constants to the log probability, and I expect that that's what you're seeing here.
from emcee.
Sorry for the confusion, I just meant that the numpy random seed were all set correctly and that this wasn't the issue.
This explanation for these small differences make a lot of sense. Thank you for your quick response!
from emcee.
Great! Yeah - I think the situation should be ok here. Feel free to re-open if related issues show up.
from emcee.
Related Issues (20)
- My samplers do not converge to maximum likelihood
- HDF backend file size
- Multiprocessing Pool scaling not as expected HOT 4
- [3.1.3] test failure with scipy 1.10 HOT 3
- Progress bar HOT 1
- Investigate the use of optree for structured parameter sets
- Issue on page /index.html HOT 1
- Add set_description to `pbar._NoOpPBar` HOT 1
- In emcee.EnsembleSampler threads command is not working
- Problem using 4-byte floats with EnsembleSampler HOT 4
- JAX implementation of emcee HOT 3
- issue with emcee installation HOT 3
- (solved) issue with sampler stalling with multiprocessing
- Saving non-arrayifiable data at each likelihood evaluation (either to disk or in memory) HOT 4
- Incompatible with numpy 2.0.0rc1 HOT 3
- New release to prepare for numpy 2 HOT 5
- Trying to use Blobs with the vectorize=True keyword returns a ValueError("incompatible input dimensions")
- emcee 3.1.5 incompatible with numpy < 1.25.0
- Using multithreading from tutorial doesnt seem to work
- Modified Walk Moves
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from emcee.