markovchain
markovchain copied to clipboard
spedygiorgio
Easy Handling Discrete Time Markov Chains
I am trying to load in and use "markovhchain" package and am getting the errors below. Does anybody have a solution please? > library(markovchain) Package: markovchain Version: 0.8.6 Date: 2021-05-17...
The hitting probabilities function is very slow for large markov chains. Users should have the option to select the target states so that unneeded values aren’t computed. Additionally, directly solving...
Is there an easy way to get the deviance from the `fitHigherOrder` function so as to compare models of different orders? For more details, see [this](https://stats.stackexchange.com/questions/147164/fit-and-evaluate-a-second-order-transition-matrix-markov-process-in-r#).
@ebbertd could you please review the tests? I was to publish the code on the CRAN but tests failed. I have added a #TODO in the testStatisticalTests.R file in the...
I know the `markovchain` package wasn't designed with Spark in mind, but I wondered if you had any ideas about how to solve a problem I am having with this...
For example, for `markovchainFit` we have the parameter `method`, which should only take values `mle`, `bootstrap`, `laplace` or `map`, but that is not checked and the following code will give...
When we call `markovchainFit` with the `laplace` or `bootstrap` method, it should warn us that the method is not available for a list. For example: ``` c1
i'm trying to fit a markov chain from a large list where each element is a customer journey between states. Each journey ends with an "end" state (so "end" is...
Hi, Could you please advise if new states occurs in chain that never ###happened before how will predict function behave ? Seems to select randomly some previous state. E.g. ;...
Hi Giorgio, I have used your R package “markovchain” on a large (>10 state) transition matrix, then merged resulting transient states for clearer presentation. I am sure there is a...