Daniel Huppmann

Oh, so you were already on the right track. Sorry that it doesn't work as expected. @peterkolp, any ideas?

@zikolach, can you take a look?

When implementing tests for this issue, this should also cover that `clone(first_model_year=y)` drops all timeseries data for `ixmp.Scenario` after (including) `y` and that the `keep_solution=False` is implemented as expected for...

**pandas** has problems with: - rows which can be cleaned used `df.dropna(how="all")` - columns that have “left-over” empty columns, naming them `Unnamed: {v}` where v is a counter… This is...

Thanks @khaeru for your (as always) very keen observations! Agree with most of them, except for two. > 5. `TimeSeries.add_timeseries()` → `add_data()`. Repeating the class name is confusing; this makes...

The problem may be that they are running concurrently - you cannot call GAMS on the same file in parallel. You would need to duplicate the `MESSAGE_run.gms` file for each...

Only the most important variables and equations are currently imported from the GAMS output to the `ixmp.Scenario` instance - these are what you'll get via `Scenario.var_list()` and `Scenario.equ_list()`. The idea...

The Java-logging will be less verbose in the next release, scheduled before the end of the summer. For the GAMS-output, that will be more tricky...

yes, #88 brings in a extended MESSAGE-scheme (v1.1) - you'll need the branch from https://github.com/iiasa/message_ix/pull/113 to be in sync between MESSAGEix and ixmp. Can you send me a screenshot of...

The problem with the checksum mismatch is due to different line endings (LF vs. CRLF) - before, that was local-system-dependent, but it will be consistent going forward. Otherwise, it would...