Problem Description
The ResultExplorer class is designed to facilitate navigation and analysis of benchmark results and generated artifacts.
Currently, ResultExplorer.load_results() returns detailed benchmark results. We want to add parameters to it so we can retrieve filtered results based on the dataset, synthesizer, or important columns.
Expected behavior
Add the following parameters to ResultExplorer.load_results():
dataset_names: list of strings for datasets; default None (show all).synthesizer_names: list of strings for synthesizers; default None (show all).- summary: boolean to display all result column or a subset of the columns, default to None. The subset of the columns is:
- Synthesizer
- Dataset
- Adjusted_Total_Time
- Adjusted_Quality_Score
- Diagnostic_Score
Additional context
- We should maintain consistency with our current practice when dealing with multiple ‘run/results.csv’ files for a given day. Specifically, if a synthesizer has been executed twice for the same dataset, then the output of
load_results() should include two separate rows.
Problem Description
The ResultExplorer class is designed to facilitate navigation and analysis of benchmark results and generated artifacts.
Currently,
ResultExplorer.load_results()returns detailed benchmark results. We want to add parameters to it so we can retrieve filtered results based on the dataset, synthesizer, or important columns.Expected behavior
Add the following parameters to
ResultExplorer.load_results():dataset_names: list of strings for datasets; default None (show all).synthesizer_names: list of strings for synthesizers; default None (show all).Additional context
load_results()should include two separate rows.