Utility methods for FIBER that do not belong into the core module. Like for visualizations, cohort filters, or time series transformation.
- Free software: MIT license
From source:
git clone https://gitlab.hpi.de/fiber/fiber-utils && \
cd fiber-utils && \
pip install -e .
- Cohort utilities
- Remove duplicated codings & filter cohorts based on information density
days_between_cohort_condition_occurrences
deduplicate_cohort
cohort_condition_occurrence_heatmap
cohort_condition_occurrence_filter
- Compare cohorts
cohort_overlap
- Time series handling
pivot_time_series
- Remove duplicated codings & filter cohorts based on information density
- Condition utilities
- Condition incidence and occurrence exploration and plotting
most_common_condition_codes_for_cohort
number_condition_occurrences_per_patient
condition_occurrence_distribution
compare_condition_incidence_in_cohort
plot_condition_first_occurrence_on_day_hist
plot_condition_last_occurrence_on_day_hist
- Quantiles & days until quantiles
condition_occurrence_quantiles_for_days
condition_occurrence_number_days_for_quantiles
- Condition incidence and occurrence exploration and plotting
- Time series utilities (symbolic and sparse symbolic representation)
sax_transform
ssr_transform
The implementation of Sparse Symbolic Representation is based on https://github.com/fbagattini/sparse_symbolic_representation. Usage granted through Creative Commons License in https://doi.org/10.1186/s12911-018-0717-4.