__init__
- BMCI.__init__(y, x, s_o)[source]
Create a QRNN instance from a given training data base y, x and measurement uncertainty given by the covariance matrix s_o.
Args
- y(numpy.array): 2D array containing the measured or simulated
brightness temperatures corresponding to the atmospheric states represented in the data base. These are sorted along the eigenvector of the observation uncertainty covariance matrix with the smallest eigentvector.
x: 1D array The retrieval quantity corresponding to the atmospheric states represented in the data base. These will be sorted along the eigenvector of the observation uncertainty covariance matrix with the smallest eigentvector.
s_o: 2D array The covariance matrix describing the measurement uncertainty.