Skip to content

Commit

Permalink
Merge pull request #106 from hubmapconsortium/thomcsmits/fix-codex-assay
Browse files Browse the repository at this point in the history 
add hth_helper to codex template
  • Loading branch information
@jpuerto-psc
jpuerto-psc authored Oct 25, 2024
2 parents 0bec054 + cd7487a commit 74e55b7
Showing 1 changed file with 10 additions and 95 deletions.
105 changes: 10 additions & 95 deletions src/user_templates_api/templates/jupyter_lab/templates/codex_clustering/template.txt
View file
Original file line number Diff line number Diff line change
Expand Up @@ -13,8 +13,8 @@
"metadata": {},
"outputs": [],
"source": [
"# !pip install --upgrade pip\n",
"# !pip install numpy pandas requests wheel matplotlib matplotlib-inline scikit-learn vitessce==3.2.6 starlette uvicorn widgetsnbextension\n"
"!pip install --upgrade pip\n",
"!pip install numpy pandas requests wheel matplotlib matplotlib-inline scikit-learn vitessce==3.2.6 starlette uvicorn widgetsnbextension hubmap-template-helper\n"
]
},
{
Expand All @@ -39,7 +39,9 @@
"from sklearn.pipeline import Pipeline\n",
"from sklearn.preprocessing import MinMaxScaler, StandardScaler\n",
"\n",
"from vitessce import VitessceChainableConfig, AnnDataWrapper"
"from vitessce import VitessceChainableConfig, AnnDataWrapper\n",
"\n",
"from hubmap_template_helper import compatibility as hth_comp"
]
},
{
Expand All @@ -59,8 +61,8 @@
"# linked datasets\n",
"uuids = {{ uuids | safe }}\n",
"\n",
"# accepted datatypes \n",
"accepted_datatypes = ['CODEX [Cytokit + SPRM]']\n",
"# accepted assay_display_names\n",
"accepted_assay_display_names = ['CODEX [Cytokit + SPRM]']\n",
"\n",
"# required filetypes\n",
"required_filetypes = ['sprm_outputs/reg001_expr.ome.tiff-cell_channel_total.csv', 'sprm_outputs/reg001_expr.ome.tiff-cell_channel_mean.csv', 'sprm_outputs/reg001_expr.ome.tiff-cell_centers.csv']\n",
Expand All @@ -73,93 +75,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"The following checks if the datasets are compatible with this template."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# This template is created for particular datatypes only.\n",
"# This functions checks for each uuids above whether they have the correct datatypes.\n",
"\n",
"def check_template_compatibility(uuids, accepted_datatypes=None, required_filetypes=None, search_api = 'https://search.api.hubmapconsortium.org/v3/portal/search'): \n",
" '''\n",
" For a set of HuBMAP UUIDs, check if valid, and return valid UUIDs.\n",
" Checks if UUIDs are present in the search API. \n",
" If accepted_datatypes is defined, checks if UUIDs are of any of the datatypes in accepted_datatypes.\n",
" If required_filetypes is defined, checks if UUIDs have all of the required filetypes in required_filetypes.\n",
"\n",
" Parameters\n",
" ----------\n",
" uuids : array of string\n",
" HuBMAP UUIDs to be checked\n",
" accepted_datatypes: array of string, optional\n",
" accepted datatypes for template\n",
" required_filetypes: array of string, optional\n",
" required datatypes for template\n",
" search_api: string, optional\n",
" URL of search API\n",
"\n",
" Returns\n",
" -------\n",
" array of string\n",
" valid UUIDs\n",
" '''\n",
" hits = json.loads(\n",
" requests.post(\n",
" search_api,\n",
" json={\n",
" 'size': 10000,\n",
" 'query': {'ids': {'values': uuids}},\n",
" '_source': ['files', 'assay_display_name']\n",
" }, \n",
" ).text\n",
" )['hits']['hits']\n",
"\n",
" # create mapping for uuid to file_types and assay_display_name\n",
" uuid_to_files = {}\n",
" uuid_to_datatypes = {}\n",
" for hit in hits:\n",
" file_paths = [file['rel_path'] for file in hit['_source']['files']]\n",
" uuid_to_files[hit['_id']] = file_paths\n",
"\n",
" hit_data_type = hit['_source']['assay_display_name']\n",
" uuid_to_datatypes[hit['_id']] = hit_data_type\n",
" \n",
" # save uuids without warnings\n",
" accepted_uuids = uuids.copy()\n",
"\n",
" # remove unvalid uuids\n",
" for uuid in uuids: \n",
" # check if all uuids are found in the search api\n",
" if uuid not in uuid_to_files.keys(): \n",
" warnings.warn('Dataset with UUID \"' + uuid + '\" not found in Search API')\n",
" accepted_uuids.remove(uuid)\n",
" continue\n",
"\n",
" if required_filetypes is not None: \n",
" # check if file_types for each uuid are in required_filetypes\n",
" file_types = uuid_to_files[uuid]\n",
" for required_file_type in required_filetypes:\n",
" if required_file_type not in file_types:\n",
" warnings.warn('Dataset with UUID \"' + uuid + '\" does not have required file type: ' + required_file_type)\n",
" if uuid in accepted_uuids:\n",
" accepted_uuids.remove(uuid)\n",
"\n",
" if accepted_datatypes is not None: \n",
" # check if assay_display_name for each uuid are in accepted_datatypes\n",
" assay_display_name = uuid_to_datatypes[uuid]\n",
" for data_type in assay_display_name:\n",
" if data_type not in accepted_datatypes: \n",
" warnings.warn('Dataset with UUID \"' + uuid + '\" has unaccepted data type: ' + data_type)\n",
" if uuid in accepted_uuids:\n",
" accepted_uuids.remove(uuid)\n",
" continue\n",
" \n",
" return accepted_uuids"
"This template is created for particular data types only. The following checks if the datasets are compatible with this template."
]
},
{
Expand All @@ -168,7 +84,7 @@
"metadata": {},
"outputs": [],
"source": [
"uuids = check_template_compatibility(uuids, accepted_datatypes=accepted_datatypes, required_filetypes=required_filetypes, search_api=search_api)"
"uuids = hth_comp.check_template_compatibility(uuids, accepted_assay_display_names=accepted_assay_display_names, required_filetypes=required_filetypes, search_api=search_api)"
]
},
{
Expand Down Expand Up @@ -705,7 +621,6 @@
{
"cell_type": "code",
"execution_count": null,
"id": "255bcd80",
"metadata": {},
"outputs": [],
"source": [
Expand All @@ -724,4 +639,4 @@
"## Then, launch the Workspace again."
]
}
]
]

0 comments on commit 74e55b7

Please sign in to comment.