Synthesize Missing Values in Numeric Data
This example shows how a distribution learned from data can be used to synthesize missing values.
Load and visualize a two-dimensional numeric dataset.
Learn the distribution underlying the data.
Plot the probability density of the distribution, along with the training data.
Use SynthesizeMissingValues to impute a missing value using the learned distribution.
By default, the imputation is done by sampling from the conditioned distribution. Obtain other possible imputations.
Visualize the position of some samples.
Visualize the distribution of imputed values.
Impute a missing value using the mode of the conditioned distribution.
Visualize the position of the mode.