feyn.__future__.contrib.diagnostics

function _get_graph_diagnostics

def _get_graph_diagnostics(
    graph
)

function print_tables

def print_tables(
    diagnostics,
    filepath=None,
    silent=False
)

Prints a list of diagnostics objects as a table. Can either print to stdout or to a file

Arguments:
    diagnostics {list(dict(str, str))} -- Your list of graph diagnostics

Keyword Arguments:
    filepath {str} -- file to save the results to (default: {None})
    silent {bool} -- whether to print the first table or not when saving to a file. (default: {False})

Raises:
    ValueError: Raises ValueError if no diagnostics are supplied, or if silent=True when writing to stdout.

function qgraph_diagnostics

def qgraph_diagnostics(
    qgraph,
    fit_df=None,
    max_graphs=None
)

Get diagnostics for your QGraph to debug errors or graph behaviour.
Use fit_df to run in-step fitting for each graph to also diagnose exceptions.

Arguments:
    qgraph {feyn.QGraph} -- Your QGraph

Keyword Arguments:
    fit_df {Union(pd.DataFrame, dict(str, np.array))} -- A dataset you'd like to debug for. Either a pandas dataframe or dict of numpy arrays (default: {None})
    max_graphs {int} -- amount of graphs to show for (mostly relevant when not using fit_df) (default: {None})

Returns:
    [list(dict(str, str))] -- a list of diagnostic objects