Hi. I was wondering if you had any insight in extending your code to include other emission models besides gaussian. In particular, how about a GMM with known number of gaussians?

I was going to take a stab at implementing it and submit a PR, but wanted to get your input first.

Thanks

Dan

What is this feature about? CD for deploying package to PyPI. It makes use of Github's workflow.

Closes Issues: https://github.com/hildensia/bayesian_changepoint_detection/issues/32

Pre-req for owner @hildensia before merging this :

1. Create a new API tokens inside your PyPI account where this project lives https://pypi.org/
2. Creating two Repository secrets inside the github project setting. (Steps defined here) A. Secret name PYPI_USERNAME which value __token__
   B. Secret name PYPI_PROD_PASSWORD with the token value from step #1

How to release a package? Leveraging github release feature This can only be done by project admin/maintainer. @hildensia Right now have made the release to based on manual action. We have to make use the of Releases option shown by github, provide a version tag number and description. If we want to change the release strategy we can update the cd.yml accordingly but usually I have seen projects follow manual release.

What Testing was done? I have tested this pipeline where the package was deployed to my Test PyPI account. https://github.com/zillow/bayesian_changepoint_detection/actions/runs/1966108484

Was removed during a PR. Is there a good way to keep cython and python in sync. I'm not sure if I prefer one over the other (python is better for debugging, cython is faster).

In the current version of code, Nw=10; ax.plot(R[Nw,Nw:-1]) is used to exhibit the changpoints. Although it works fine, I am really confused about the moral behind it. I tried to plot the run length with maximum prob in each time step i.e. the y index of maximum prob in each x col, but the result showed the run length keeps going up... I also went back to Admas's paper but found nothing about change point indentification stuff (he just stop at R matrix)... I also tried to find Adams's MATLAB code, but the code seems to have been removed...

I am trying to use this method in my work, and I believe it's the best to fully understand it before any deployment. Any help will be appreciated and thanks a lot!

The scale factor should be the standard deviation. There was a small bug in the betaT0 calculation, this makes it consistent with the paper/gaussdemo.m file.

If I click on the "example notebook" work - an nbviewer link - I get a "too many redirects" error.

It would be nice if the example notebook was easily accessible in the repo (maybe I overlooked it... ) because we don't need a live notebook / nbviewer to figure out whether the example fits our use case.

I think my question is related to the one, which was not answered and is already closed: https://github.com/hildensia/bayesian_changepoint_detection/issues/19

In your example, you have applied the student t-distribution as a likelihood. I understand the distribution, its parameters, but I have a question about how you set up prior and update its parameters in the code. So the following is:

df = 2*self.alpha
scale = np.sqrt(self.beta * (self.kappa+1) / (self.alpha * self.kappa))

I don't understand what alpha, beta and kappa correspond to. How have you come across this expression? The paper by Adams and McKey refers to updating sufficient statistics. Is your expression related to that? If so, how can I do that for any other distribution, let's say gaussian? In my comment, I refer to the following formula in the paper:

Hi guys,

there is an import issue if one uses newer scipy versions.

Would be a quick fix if you adapt the import statement at offline_changepoint_detection.py

try:  # SciPy >= 0.19
    from scipy.special import comb, logsumexp
except ImportError:
    from scipy.misc import comb, logsumexp  # noqa

• Introduces a pluggable prior/posterior config for multivariate Gaussian data, with sensible defaults. Note that this only works for scipy > 1.6.0, where they introduced the multivariate t PDF. The library will remind you to upgrade if you have an old version.
• Adds a test for this new configuration, as well as for the univariate one
• Adds a "dev" and "multivariate" setup extra, meaning that you can pip install bayesian_changepoint_detection[dev] for development work (currently this installs pytest), or pip install bayesian_changepoint_detection[multivariate] (enforces that you have a new enough scipy version for this new feature)

I ran the given online detection example in the notebook, and I assumed the y axis indicating the probability of changepoint (am I right?). But the y value ranged from zero to hundreds. I am not very familiar with the math, so can anyone please explain this outcome?

Thanks.

This fixes the full cov method and adds an example similar to the original ipython notebook. If you prefer, I can merge them separately, but since they are related, I thought it'd be fine to merge them together.

Hi，nice to meet you，and i want to aks a basic question，if i don’t know the distribution of data（not the normal distribution），then could i use the bocpd？ Thank you！

Hi, I've noticed is the scaling of the data can have an effect on the result, but I am not sure why it would and can't find any reason for it in the code or references. Below I have the CP probabilities for the same data with or without a constant factor, which are somewhat different.

Are there some assumptions about the input data I am missing? Thanks

There is always a tradeoff between false alarms and missed alarms, and when the algorithm is more sensitive we should have higher false alarm rate and lower missed alarm rate. My question is, is it possible to adjust the sensitivity level of this algorithm by changing the hyperparameter (e.g., alpha, beta, kappa, mu)? Thank you!

Hi,

I installed bayesian_changepoint_detection from this github repository.

By setting (accidentally) np.seterr(all='raise'), I was able to cause the following exception.

I am not sure whether this would have any relevance for the further processing, but I just wanted to draw attention to people working on / with this library.

/home/user/venv/env01/bin/python3.6 /home/user/PycharmProjects/project01/snippet.py
Use scipy logsumexp().
Traceback (most recent call last):
  File "/home/user/PycharmProjects/project01/snippet.py", line 68, in <module>
    Q, P, Pcp = offcd.offline_changepoint_detection(data, partial(offcd.const_prior, l=(len(data) + 1)), offcd.gaussian_obs_log_likelihood, truncate=-40)
  File "/home/user/experiments/original-unforked/bayesian_changepoint_detection/bayesian_changepoint_detection/offline_changepoint_detection.py", line 98, in offline_changepoint_detection
    Q[t] = np.logaddexp(P_next_cp, P[t, n-1] + antiG)
FloatingPointError: underflow encountered in logaddexp

Process finished with exit code 1

For someone whoever is interested, I have added Normal known precision, poisson distributions in my fork below. Also tried adding another type of hazard function which is normally distributed over time. Usage of the same is updated in Example code as well. Find my fork here - https://github.com/kmsravindra/bayesian_changepoint_detection

Hi,

I am confused about the returned R matrix interpretation in the online detection algorithm. In the notebook example, the third plot is R[Nw,Nw:-1], where it is mentioned to be "the probability at each time step for a sequence length of 0, i.e. the probability of the current time step to be a changepoint." So why do we choose the indices R[Nw,Nw:-1] ? why not R[Nw,:]

Also, it was mentioned as an example that R[7,3] means the probability at time step 7 taking a sequence of length 3, so does R[Nw,Nw:-1] means that we are taking all the probabilities at time step Nw ?

Any suggestions to help me to understand the output R ?

Thanks