| Age | Commit message (Collapse) | Author |
|---|
|
|
|
- Added all current SNP browser code (not complete yet)
- Added change to convert_geno_to_bimbam that makes it ignore .geno files marked as "filler" (so ones where the .geno file is fake and we sometimes directly receive the genotypes as BIMBAM)
- Changes TheSpecies object in species.py to accept species name as well as dataset name
|
|
|
|
|
|
just clicking their row in collection
- Cofactor color picker now works in Safari/Macs
- Displays N for relevant samples in trait page sample table
- Don't show bar chart when N>256
- Mapping loading page contents better centered
- Anonymous collections timeout correctly listed as 30 days now
- Minor allele frequency can actually be changed for GEMMA now (previously didn't work)
- Fixed transcript position marker location for mapping results
- Notifies user if their e-mail isn't associated with an account when they attempt to request forgotten password
- Users can now map with submitted traits
- Histogram width changes depending upon number of bins (need to improve this still)
- Improved Q-q plot (previously called "probability plot")
|
|
Gemma optimized for production
|
|
|
|
|
|
|
|
|
|
Added "change user to elasticsearch" to the description of how to start elasticsearch
|
|
|
|
|
|
|
|
|
|
Added Phenogen track to mapping results
Added comparison bar chart figure
Simplified global search to not build trait/dataset objects, which speeds thing up considerably
Fixed correlation matrix to correctly deal with 0 values
Fixed issue where anonymous collections couldn't be created if none already existed
|
|
I added the command to start elasticsearch, since I've had to look it up/ask Pjotr about it a couple times now.
|
|
Figures now correctly rescale when transforms are applied
Fixed height of probability plot
Removed some text related to pylmm since it's currently not being used
|
|
trait page
Got figures mostly working with transformed values, except for changing the y-axis range of one
Fixed issue with the home page "Make Default" button position (as well as the drop-downs in general)
Added timeout to Elasticsearch connection to fix related error
|
|
previously
|
|
should have their full functionality back (adding/removing collections, adding/removing traits, etc)
|
|
staging, though still not sure why it's working on production without that change
Added script to convert the dryad format genotype files to BIMBAM
removed db_uri from parameters of parse_db_uri in gen_select_dataset.py, since it can now just pull it from settings as a global variable
|
|
sample data into ElasticSearch
Added option to replace trait page sample/strain values with normalized ones
Began editing Lei's scatterplot code
Changed elasticsearch_tools' get_elasticsearch_connection so that it can also be used for purposes other than user authentication (by adding a "for_user" parameter)
|
|
|
|
|
|
Fixed issue where GEMMA + LOCO wasn't working with AIL (and probably some other groups)
|
|
instead of -lmm 1 (Wald test)
Added script to convert .geno files to JSON to maintenance folder (geno_to_json.py)
|
|
set width
Changed to using BIMBAM files to build Marker objects in dataset.py
|
|
|
|
some problem. Probably better to leave this file alone for the time being.
|
|
|
|
|
|
Removed some unused code from the other files listed
|
|
marker_regression_gn1.py
marker_regression.py
do_search.py
export_traits.py
heatmap.py
GeneUtil.py
heatmap.html
|
|
Testing
|
|
testing
|
|
Testing
|
|
* I can't swear on this, but it seems the presence of these statements
was causing elasticsearch to act weird.
|
|
|
|
|
|
* Update the indexes mappings to be compatible with the newer
Elasticsearch 6.2.* series.
Close the index before updating it, and reopen it after to help with
the re-indexing of the data.
* Update the error logger to include the exception that was thrown.
|
|
|
|
* Prevent elasticsearch from analysing and tokenising the email_address
field so as to avoid issue with getting back all email addresses with
the same domain as the one being searched for.
|
|
* Add tests to ensure the "Forgot Password" feature works appropriately.
|
|
* Use existing code to get the elasticsearch connection. This should
prevent tests from failing in case the way connections to
elasticsearch are made change.
* Delay a while after deleting to allow elasticsearch to re-index the
data, thus preventing subtle bugs in the test.
|
|
* Ensure the return is at the correct indentation level so that it is
executed correctly.
|
|
* Ensure that the user actually provides an email address when they try
to use the forgot password feature.
|
|
* Use the LOG_LEVEL provided by the application settings to enable easy
control of logging when running tests. It helps avoid a lot of
verbosity when running tests.
|
|
* Remove the run-integration-tests.py, and use the test-website.py as
the entry point for all tests. This simplifies running the tests for
the site.
|
|
|
