diff options
Diffstat (limited to 'gn3/db/sample_data.py')
-rw-r--r-- | gn3/db/sample_data.py | 38 |
1 files changed, 15 insertions, 23 deletions
diff --git a/gn3/db/sample_data.py b/gn3/db/sample_data.py index b67fe1c..96168f1 100644 --- a/gn3/db/sample_data.py +++ b/gn3/db/sample_data.py @@ -63,29 +63,21 @@ def get_trait_csv_sample_data( conn: Any, trait_name: int, phenotype_id: int ) -> str: """Fetch a trait and return it as a csv string""" - __query = ( - "SELECT concat(st.Name, ',', ifnull(pd.value, 'x'), ',', " - "ifnull(ps.error, 'x'), ',', ifnull(ns.count, 'x')) as 'Data' " - ",ifnull(ca.Name, 'x') as 'CaseAttr', " - "ifnull(cxref.value, 'x') as 'Value' " - "FROM PublishFreeze pf " - "JOIN PublishXRef px ON px.InbredSetId = pf.InbredSetId " - "JOIN PublishData pd ON pd.Id = px.DataId " - "JOIN Strain st ON pd.StrainId = st.Id " - "LEFT JOIN PublishSE ps ON ps.DataId = pd.Id " - "AND ps.StrainId = pd.StrainId " - "LEFT JOIN NStrain ns ON ns.DataId = pd.Id " - "AND ns.StrainId = pd.StrainId " - "LEFT JOIN CaseAttributeXRefNew cxref ON " - "(cxref.InbredSetId = px.InbredSetId AND " - "cxref.StrainId = st.Id) " - "LEFT JOIN CaseAttribute ca ON ca.Id = cxref.CaseAttributeId " - "WHERE px.Id = %s AND px.PhenotypeId = %s ORDER BY st.Name" - ) case_attr_columns = set() csv_data: Dict = {} with conn.cursor() as cursor: - cursor.execute(__query, (trait_name, phenotype_id)) + cursor.execute(""" +SELECT concat(st.Name, ',', ifnull(pd.value, 'x'), ',', +ifnull(ps.error, 'x'), ',', ifnull(ns.count, 'x')) AS 'Data', +ifnull(ca.Name, 'x') AS 'CaseAttr', ifnull(cxref.value, 'x') AS 'Value' +FROM PublishFreeze pf JOIN PublishXRef px ON px.InbredSetId = pf.InbredSetId +JOIN PublishData pd ON pd.Id = px.DataId JOIN Strain st ON pd.StrainId = st.Id +LEFT JOIN PublishSE ps ON ps.DataId = pd.Id AND ps.StrainId = pd.StrainId +LEFT JOIN NStrain ns ON ns.DataId = pd.Id AND ns.StrainId = pd.StrainId +LEFT JOIN CaseAttributeXRefNew cxref ON (cxref.InbredSetId = px.InbredSetId +AND cxref.StrainId = st.Id) LEFT JOIN CaseAttribute ca +ON ca.Id = cxref.CaseAttributeId WHERE px.Id = %s +AND px.PhenotypeId = %s ORDER BY st.Name""", (trait_name, phenotype_id)) for data in cursor.fetchall(): if data[1] == "x": csv_data[data[0]] = None @@ -102,11 +94,11 @@ def get_trait_csv_sample_data( return "Strain Name,Value,SE,Count\n" + "\n".join(csv_data.keys()) columns = sorted(case_attr_columns) csv = "Strain Name,Value,SE,Count," + ",".join(columns) + "\n" - for key, value in csv_data.items(): - if not value: + for key, _value in csv_data.items(): + if not _value: csv += key + (len(case_attr_columns) * ",x") + "\n" else: - vals = [str(value.get(column, "x")) for column in columns] + vals = [str(_value.get(column, "x")) for column in columns] csv += key + "," + ",".join(vals) + "\n" return csv return "No Sample Data Found" |