GMMfit: Fit a Gaussian model to a set of channels
In ArtemSokolov/naivestates: Cell state calling using Naive Bayes

Fit a Gaussian model to a set of channels

1	GMMfit(X, cid, ..., bounds, baseline = 0.01, mu_init = c(0.2, 0.8), seed = 100)

`X`	- cell-by-channel data frame of marker expression
`cid`	- column name or index that contains cell IDs
`...`	- columns to fit a GMM to
`bounds`	- [optional] Two values specifying the range of values to be used for mixture modeling. This allows for exclusion of outliers from model fitting. Calculated automatically, if not provided.
`baseline`	- [optional] baseline for automatic boundary calculation
`mu_init`	- [optional] vector of two values, each between 0 and 1, specifying the initial placement of Gaussian means.
`seed`	- [optional] random seed to allow for reproducibility