Tests should be added to verify that bootstrap_dea() returns correct results. In order to achieve this, we need to identify data sets with known properties for bias corrected efficiency estimates and confidence intervals. Note that adding tests for bootstrap will incur a performance hit on the test suite, as we need a large number of iterations in order get reproducible results (down to fourth decimal), as we are using (by definition) a stochastic function inside the bootstrap.
Tests should be added to verify that
bootstrap_dea()returns correct results. In order to achieve this, we need to identify data sets with known properties for bias corrected efficiency estimates and confidence intervals. Note that adding tests for bootstrap will incur a performance hit on the test suite, as we need a large number of iterations in order get reproducible results (down to fourth decimal), as we are using (by definition) a stochastic function inside the bootstrap.