The origin Mapper includes mean and majority label embeddings. And this function provides another way to color the Mapper nodes. The function is useful to connect original data for color labeling, especially if you're interested in characteristic attributes.
CPEmbedding(
mapper,
original_data,
columns = list(),
a_level = NULL,
b_level = NULL
)A list of conditional probabilities value for each Mapper node.
A Mapper object created by the MapperAlgo function.
Original dataframe, not the filter values.
Two columns in original_data to compute conditional probability.
The level (attribute) of column A to condition on. If NULL, the first level is used.
The level (attribute) of column B for which the conditional probability is computed. If NULL, the first level is used.