There are files with names beginning with rtest located in the tests folders (see discussion #1047). What they do is load up the R version of chainladder and then test chainladder-python's functions against them to check for parity.
However, these files are not run during testing and negatively impact test coverage.
It's been proposed that we run the R code locally to acquire static expectation values to test against. We should make a note of what R version was used to do this
| File Path |
chainladder/methods/tests/rtest_mack.py |
chainladder/core/tests/rtest_correlation.py |
chainladder/development/tests/rtest_clark.py |
chainladder/development/tests/rtest_munich.py |
chainladder/development/tests/rtest_development.py |
chainladder/tails/tests/rtest_exponential.py |
These files haven't been touched in 4 years, so it's unknown to me whether they work or if they worked the last time they were edited. I also propose that if we find this task to be unmanageable, we could move the files to their own branch outside of main.
There are files with names beginning with
rtestlocated in the tests folders (see discussion #1047). What they do is load up the R version of chainladder and then test chainladder-python's functions against them to check for parity.However, these files are not run during testing and negatively impact test coverage.
It's been proposed that we run the R code locally to acquire static expectation values to test against. We should make a note of what R version was used to do this
chainladder/methods/tests/rtest_mack.pychainladder/core/tests/rtest_correlation.pychainladder/development/tests/rtest_clark.pychainladder/development/tests/rtest_munich.pychainladder/development/tests/rtest_development.pychainladder/tails/tests/rtest_exponential.pyThese files haven't been touched in 4 years, so it's unknown to me whether they work or if they worked the last time they were edited. I also propose that if we find this task to be unmanageable, we could move the files to their own branch outside of main.