generate_weather: Make predictions using calibrated models and a new data frame
In leerichardson/sdsmR: Implementation of Statistical Downscaling

This function is meant to be used after running the calibrate_models function. The idea is that it checks the model type, and then works through the different subsets of the new dataframe to make predictions, and returnsa data frame with dates, and predictions, and ensembles of predictions if that option has been specified.

1 2	generate_weather(models, new_dataframe, uncertainty = "ensemble", num_ensembles = 1, y = NULL)

`models`	A list containing either 1, 4, or 12 models depending on whether the model is annual, seasonal, or monthly.
`new_dataframe`	A new dataframe in which predictions are to be made from the calibrated models.
`uncertainty`	Either "ensemble" or "interval".
`num_ensembles`	If uncertainty = "ensembles", then this specifies how many ensembles to return. Default is set to one.
`If`	sing an autoregressive model, you must specify the response variable.