ts_forecast.array: Calculate a timeseries forecast from an array
In kdayday/forecasting: Probabilistic Solar Forecsting

Construct a time series of power forecasts, using input training data. Assumes training data already captures differences in magnitude (i.e., power rating) amongst sites. Forecast is NA for times when sun is down.

1
2
3

## S3 method for class 'array'
ts_forecast(x, start_time, time_step, scale, location,
  method, sun_up_threshold = 0.5, MoreTSArgs = NA, ...)

`x`	An array of training data of dimensions [time x ntrain x nsites] for n-dimensional forecast
`start_time`	A lubridate time stamp
`time_step`	Time step in hours
`scale`	One of 'region', 'total'
`location`	A string
`method`	One of 'gaussian', 'empirical', 'vine' (irrelevant if scale == 'site')
`sun_up_threshold`	An absolute [MW] threshold on the ensemble members to remove dubious sunrise/sunset valud
`MoreTSArgs`	An optional dictionary of time-series arguments to the forecast calculation
`...`	optional arguments to the prob_forecast object