Marcel Rieger
Marcel Rieger
We should have generic tasks within columnflow that produce plots that don't show a variable distribution in a single category, but that show several categories at once. I see two...
It looks like conda-pack does not change references to executables when creating the bundle of our basic software environment. This results in some executables (like micromamba itself) to be present...
The values at ticks on the x axis for variables that have `discrete_x` defined are currently just derived from `bin_edges` which can lead to too many ticks being shown. This...
The new analysis template mechanism would allow for running tests of the entire *standard* workflow, from LFN retrieval down to plots. We should create a CI job that runs on...
As the project is evolving towards its beta phase, the packaging should be updated such that it would be no longer required to install it as a submodule of an...
We define our statistical models in an experiment agnostic way, allowing for various formats for exporting actual fit models (stat. model + data). So far, we have a combine datacard...
#145 added a first simple separation CMS-related and generic analysis code in terms of tasks and producers/calibrators. After some deeper dive, there is still a handful of parts that exhibit...
When creating a plot with `cf.PlotShiftedVariables1D`, the ratio pad is comparing the up/down shifted histograms with the nominal one  The two ratio lines are shown as histograms which...
While #106 will provide an entry point for new users, we should also plan additional documentation for our [http://columnflow.readthedocs.io](rtd) page. We can use *this* issue to assemble some topics for...
Hi! I noticed that, when specifying the language of a code block as in the following ```bash $ git lfs install Updated git hooks. Git LFS initialized. ``` the actual...