allow bootstrapping for time series

[olivierelew]

Hey,

There needs to be a way to add different or custom bootstrapping methods to the function "bootstrap_model()". Especially when dealing with time series data does the regular method create some concern.

source: "Marketing Mix Modeling Using PLS-SEM, Bootstrapping the Model Coefficients"
author: "Mariano Méndez-Suárez"

quote: This research shows the risks of using in sampling bootstrap (regular approach) for time series data because the method destroys the internal structure of the series and shows wider confidence intervals for the outer loadings of the models.

solution: This study recommends using meboot bootstrapping as an alternative and proves its suitability
for time series or marketing mix modeling with PLS-SEM.

explanation meboot: "The meboot algorithm [1] is a procedure that generates a large number of replicates,
e.g., 5000, of the original series, which can be used for statistical inference; it then applies
the “blocking” technique to break the time series into nonoverlapping blocks such that
the grand mean of all the simulated samples equals the time average of the original,
constructing bootstrap samples, or ensembles, that retain the basic shape and dependence
structure of the original data."
[1] Vinod, H.D.; López-de-Lacalle, J. Maximum Entropy Bootstrap for Time Series: The meboot R Package. J. Stat. Softw. 2009, 29,
1–29. [CrossRef]

Could you look at the recommendations of the research done above and look at different bootstrapping methods which can be added as a functionality? In particular the meboot approach for better bootstrapping of time series data.

*Quotes are taken/derived from the research paper.

[NicholasDanks]

Dear @olivierelew this is a great idea. If you could implement it in the code and make a pull request, I would be very happy to work with you on implementing it.