pydlm
pydlm copied to clipboard
wwrechard
A python library for Bayesian time series modeling
All my plots with plotpredictN seems to start with the first predicted point above or below the point selected by date= . To clarify what I mean, I think the...
Hi, Does the package account for any feature normalization when variables of different scales are passed into the model? Square root matrix normalization or something similar? If not, is it...
Hello, I'm attempting to using binary data in the PYDLM algorithm, similar to BSTS (R). The model fits, but it returns "nan" for predicted values. Does PyDLM not accept binary...
Hello, I've used this package with some success to model hundreds of time series so thanks for the effort ! I'd like to use it in production but I've run...
first, I just want to thank you for building this awesome library. Very impressive and extremely useful. I'm running into a bug where I cannot append new data to a...
Hi, I'm trying to retrieve the raw data used in plotCoef. I'm in the idea that the red line is taken via .getLatentState(filterType='backwardSmoother', name='x') but my issue is that I...
It's clear for me why .fit() time complexity (execution time) depends on time series length. Probably O(N^2). But why it's true for predictN(n=1) as I found out ? It looks...
Hi, I was searching for an alternative for bsts(R) in Python and stumbled on this gem. As far as I understand there is no difference between bsts and pydlm, except...
Hi, Is there any way to get separate components for future prediction, let's say I just want to get trend component of future predictions.
Is there any plan to extend the module to allow for dynamic factor models where latent states are fewer than observed time series?
