- Fixed a bug with individual plots when using LFMM (which would cause the wrong covariate to be shown in the plot). Another big thank you to Gabriele Nocchi for finding it!
- You can now provide a file with covariate names (one per line), and all plots will use this information instead of simply numbering the covariates.
- If you do not provide this file, by default your covariates will be numbered, starting with 1 (previous to version 0.4.4 they were 0 indexed).
- Fixed a bug with delimiters on LFMM input files (files using tab as a delimiter would cause python tracebacks). A big thank you to Gabriele Nocchi for finding it!
- Corrected a switched "Latitude" and "Longitude" information in the manual. Once more, big thank you Gabriele Nocchi for spotting it and reporting!
- Minor improvements to LFMM2_Workflow.R
- Corrects a bug introduced by changes in numpy 1.19.0
- Adds a new workflow script to be able to use lfmm2
- Changes the Bioconductor install method to the current one
- Adds infrastructure to use Gitlab's CI server parallel to Travis CI.
- Fixed a broken help link
- Adds a new feature: RONA maps
- To use it, just add the option
-mapwith a path to where you want to save the map, such as~/my_RONA_map.png - In order to use this feature, the dependency cartopy is required.
- The manual now contains instructions on installing
cartopydepending on your OS- Instructions for OSX are still missing. Help, anyone?
- For now, the options assumes that the first variable in the environmental file is "Latitude" and the second one is "Longitude"
- To use it, just add the option
- Immutables option is no longer on by default
- Improved help text of the immutables option
- Improved handling of LFMM population data (as opposed to individual data)
- Tests altered to conform to new options
- Better handling of individual plots
- Better PEP8 conformance
- Corrects a bug in when using outliers
- Better PEP8 conformance
- Implements automated test coverage measurements
- Corrects a bug in the installation of the R script
LFMM_workflow.R - Better PEP8 conformance
pyRonacan now use LFMM results as an input to calculate RONA- The test suite was also increased to match the changes
- Major code re-organization
- Can now be installed via
setup.py - Pypi submission
- Adds unit tests
- Travis-CI integration
- Proper documentation is started
- Autosave plots instead of interactive (which allows for a lot better automation)
- Bugfix with
arg.immutablesargument.
- Better PEP8 conformance
- Weighted R² means in the summary (when requested)
- Better label handling in the plots
- Option to choose which covariates are immutable and skip them
- Added a new line to the summary (Number of SNPs associated with each covariate)
- Can now parse multiple outputs from BayPass and not only for the MCMC_core model.
First fully working version. There is still a lot to do to turn the program into a true "hit", but the basics are in.