Lisflood Code Versions Save

NetCDF reader for forcings (or any other temporal data) now based on Xarray, allowing more flexibility and efficiency (dramatic improvement in calibration mode, i.e. without NetCDF outputs). Two new options in the xml settings file: NetCDFTimeChunks: chunking size in the time dimension. Recommended value is “auto" but chunking size can be specified manually or set to “-1" to load the whole time series into memory (very fast but expensive in terms of memory). MapsCaching (True or False): option designed for the lisflood calibration. If set to True, all the static maps and forcings will be stored in a cache so that they don't have to be loaded by each lisflood instance. This option sets the value of NetCDFTimeChunks to "-1", meaning that the whole time series in the NetCDF inputs is loaded into memory. Several tests have been added: lat/lon domains, inflows, new reader, etc. This version also makes it easier to use lisflood as a library and to run multiple instances of lisflood in a thread-safe environment. This can be done by doing the following: import lisf1 ... lisf1.main(settings_file, flags) …

• Fix import order in unit tests to avoid name clash: for a new installation, installing an old verison of lisflood-utilities package and its dependencies (GDAL, netCDF)  was making PCRaster libs to segfault during tests.
• Update tox and pytest configurations, add custom tox ini file to gitignore.
• Improve setup script to install correct GDAL python package
• Updated Docker image:

https://hub.docker.com/repository/docker/efas/lisflood

This release includes:

• fix for High Resolution grids (thanks @agahirpa)
• requirements files with fixed versions to enforce reproducible dev environments
• longer simulation tests (from 1month to 6 months)
• checking TSS output as well

All unit tests passed (py27, py36, py37) All black box tests passed (see https://github.com/ec-jrc/lisflood-tests for more info)

This is a minor release with enriched unit tests suite.

This release is verified and tested with https://github.com/domeniconappo/lisflood-tests test suite.

Test suite: EFAS daily 2,5 years EFAS 6hourly 1,5 years GLOFAS daily 1,5 years

Reference version as agreed: https://github.com/ec-jrc/lisflood-code/commits/92061c80f2613af89f0d03cd1d2a0ccf3171a1f8

Reference data were executed with Python2.

NOTE: When executing tests with latest versions, we had to re-introduce - in the version under test - a bug that was present in reference version (ie a division that was casted to integer in Python2). This was necessary to be sure that python3 changes and mass refactoring didn't touch the mathematical model. Of course, we removed the bug.