NEWS.md
all.ss
in function get_stable_state_from_models_dir()
to include all stable states in the returned tibble
objectget_stable_state_from_models_dir()
(return empty tibble when no models had stable state).gitsbe
files, now disregard other kind of files that might be inside these directories.DESCRIPTION
fileupdate_biomarker_files
function (writes to tmpdir()
instead of the user’s library directory)NaN
MCC scores were produced (which is the correct limiting value - see Chicco at al. (2020)). Thus, the previous versions handling of NaN
MCC scores, is now deprecated.penalty
parameter to account for the difference in model group size when calculating the average activity or link operator data differences. This minimizes the bias in the returned biomarkers.
emba::get_vector_diff
and the corresponding StackOverflow question.penalty=0
in the general emba::biomarker_*
functions (though the results will probably be very biased and that’s why the default value for the penalty
is now 0.1).models.stable.state
parameter used in various functions can take any values in the [0,1] interval and not just 0 (inactive) and 1 (active).models
anymore:
data.frame
object instead of a matrix
..gitsbe
extension anymore.emba::get_link_operators_from_models_dir
, emba::get_stable_state_from_models_dir
and emba::get_model_names
.emba::biomarker_mcc_analysis
and emba::biomarker_tp_analysis
do not use the calculate.subsets.stats
input option anymore. The emba::biomarker_synergy_analysis
continues to do so and now also calculates and returns all possible synergy set and subset pairs that miss just one of the model predicted synergies (emba::get_synergy_comparison_sets
).get_synergy_scores
now supports reading both ensemble-wise and model-wise synergies filescalculate.subsets.stats
option to the general analysis functions (biomarker_*
) that decides if the powerset of the observed synergies and the number of models predicting each subset is going to be calculated. The default value is set to FALSE
to save computation time :)get_synergy_scores
get_avg_link_operator_diff_based_on_synergy_set_cmp
get_avg_link_operator_diff_based_on_specific_synergy_prediction
filter_network
- to use for visualizing induced subgraphsusefun
min version to 0.4.3)count_models_that_predict_synergies
function and added tests for it. For a benchmark see relative Stack Overflow thread