lineage.crossval
Cross validation.
Attributes
Functions
|
Taking a list of lineages and the percentage of cells want to be masked, it marks those x% negative. |
|
Perform cross validation for the experimental data which runs in parallel for all states. |
|
Given the complete population, it masks 25% of cells and prepares the data for parallel fitting using crossval function. |
Module Contents
- lineage.crossval.exe
- lineage.crossval.hide_observation(lineages, percentage, rng=None)
Taking a list of lineages and the percentage of cells want to be masked, it marks those x% negative.
- Parameters:
lineages (list)
percentage (float)
- Return type:
list
- lineage.crossval.crossval(train_populations, num_states, rng=None)
Perform cross validation for the experimental data which runs in parallel for all states. :param train_populations: the populations after applying hide_observation. This includes the list of list of lineages. :param hidden_indexes: is a list of list of np.arrays for each lineage, filled with zeros and ones. ones refer to the index of those cells that have been hidden. :param hidden_obs: list of list of tuples of observations that have been masked in the train_lineage. :param num_states: is a range of states we want to run the cross validation for.
- Parameters:
train_populations (list)
num_states (numpy.ndarray)
- lineage.crossval.output_LL(complete_population, desired_num_states)
Given the complete population, it masks 25% of cells and prepares the data for parallel fitting using crossval function.