Christoph Knote (5ab24752) at 01 May 06:47
Update README.md
Christoph Knote (98259dee) at 01 May 06:46
Update README.md
Christoph Knote (1fc67769) at 28 Apr 15:38
Ready for primetime
The title says it all - anything that needs fixing before we release, @FObersteiner ? Please create additional issues with milestone publish as v2.0.0 on pypi. Thanks!
Various small updates to make package ready to be released to PyPi.
Various small updates to make package ready to be released to PyPi.
Christoph Knote (ec21b3b7) at 28 Apr 11:43
documentation upload: use masked variable to export base64 encoded ...
This merge requests adds a create_docs
task to our .gitlab-ci.yml
. The task will generate docs using sphinx and then push the results to our web server using rsync.
The job uses 3 variables, but only two of them are exposed in the CI configuration:
create_docs:
(...)
variables:
docs_upload_host: hosted-024-173.rz.uni-augsburg.de # hostname of target web server
docs_path_component: icartt # if your path/URL should contain a stub, different than the (Gitlab) project name.
The third variable is docs_ssh_key
which is private an thus has to be configured as a masked variable
using the Gitlab Web UI.
The proposed task fetches the current projects version using poetry version -s
, e.g. for the current HEAD it would upload to: https://mbees.med.uni-augsburg.de/docs/icartt/2.0.0-rc1/
You may find some internal documentation on how to set it up on https://mbees.med.uni-augsburg.de/dokuwiki/doku.php?id=it:services:documentationhosting
This merge requests adds a create_docs
task to our .gitlab-ci.yml
. The task will generate docs using sphinx and then push the results to our web server using rsync.
The job uses 3 variables, but only two of them are exposed in the CI configuration:
create_docs:
(...)
variables:
docs_upload_host: hosted-024-173.rz.uni-augsburg.de # hostname of target web server
docs_path_component: icartt # if your path/URL should contain a stub, different than the (Gitlab) project name.
The third variable is docs_ssh_key
which is private an thus has to be configured as a masked variable
using the Gitlab Web UI.
The proposed task fetches the current projects version using poetry version -s
, e.g. for the current HEAD it would upload to: https://mbees.med.uni-augsburg.de/docs/icartt/2.0.0-rc1/
You may find some internal documentation on how to set it up on https://mbees.med.uni-augsburg.de/dokuwiki/doku.php?id=it:services:documentationhosting
Christoph Knote (500c6bf9) at 19 Apr 12:34
Fix Python dep
Christoph Knote (91e7a423) at 19 Apr 12:32
Non-ASCII character slipped through.
Christoph Knote (ba5c5708) at 19 Apr 12:28
Minor updates
__str__
methods for the dataset and variable classes. I like to use those to get a quick insight of what's in the dataset / variable.I can make a draft soon.
in the dataset's __init__
,
self.version = None # should this be 2.0 by default, for ICARTTv2?
and directly on-topic,
if not self.isValidFileName(pathlib.Path(f).name):
warnings.warn(f"{pathlib.Path(f).name} is not a valid ICARTT filename")
# else -> split on "_", then set first part as dataID, second part locationID?
In this context, I also noted that dataset.revision
is not updated if reading an ict file. If the revision is not 0, makeFileName()
will return an incorrect result.
missing values can be specified as an integer, however if the data contains missing values, they will be written as floats (or what ever the user specifies.
the world was easy when there were just integers in the data and floating point scaling factors ;-)
Sames is true for the data interval code btw. - if you read a file where it is specified as integer, and write that same file back to disc with the write method, the interval code will be specified with a decimal point (e.g. in: 0
, out: 0.0
).