check.windowsize: Bootstrap convergence check with subsampling
In tbonne/netTS: netTS: network time series

This function will estimate the convergence of the chosen network measure using bootstrapped samples of the data, while also checking for measurement sensitivity to data subsampling.

check.windowsize(data, windowsize = days(30), windowshift = days(1),
  directed = FALSE, measureFun = degree, corFun = 1, boot.samples = 100,
  SRI = FALSE, probs = c(0.025, 0.975), subsamples = c(1, 0.8, 0.6),
  plot = TRUE)

`data`	Dataframe with relational data in the first two rows, and a time stamp in the third row. Note: time stamps should be in ymd or ymd_hms format. The lubridate package can be very helpful in organizing times.
`windowsize`	The size of each window in which to generate a network.
`windowshift`	The amount of time to shift the window when generating networks.
`directed`	Whether to consider the network as directed or not (TRUE/FALSE).
`measureFun`	The measurment function to be used with the bootstapped networks.
`corFun`	The method used to compare observed values with bootstrapped values: 1-Cosine similarity, 2-pearsons correlation, 3-Euclidean distance
`boot.samples`	The number of bootstrapped samples to run (Default=100)
`SRI`	Wether to use the simple ratio index (Default=FALSE)
`probs`	The quantiles of the bootrap samples to return (Default=c(0.025,0.975)).
`subsamples`	A vector of values between 0-1 used to subsample the original dataframe.
`plot`	Wether a plot of the results should be produced.