aboutsummaryrefslogtreecommitdiff
path: root/wqflask/base/data_set.py
diff options
context:
space:
mode:
Diffstat (limited to 'wqflask/base/data_set.py')
-rw-r--r--wqflask/base/data_set.py51
1 files changed, 47 insertions, 4 deletions
diff --git a/wqflask/base/data_set.py b/wqflask/base/data_set.py
index 178234fe..468c4da0 100644
--- a/wqflask/base/data_set.py
+++ b/wqflask/base/data_set.py
@@ -115,7 +115,8 @@ Publish or ProbeSet. E.g.
except:
pass
- self.redis_instance.set("dataset_structure", json.dumps(self.datasets))
+ self.redis_instance.set(
+ "dataset_structure", json.dumps(self.datasets))
def set_dataset_key(self, t, name):
"""If name is not in the object's dataset dictionary, set it, and update
@@ -154,10 +155,12 @@ Publish or ProbeSet. E.g.
if t in ['pheno', 'other_pheno']:
group_name = name.replace("Publish", "")
- results = g.db.execute(sql_query_mapping[t].format(group_name)).fetchone()
+ results = g.db.execute(
+ sql_query_mapping[t].format(group_name)).fetchone()
if results:
self.datasets[name] = dataset_name_mapping[t]
- self.redis_instance.set("dataset_structure", json.dumps(self.datasets))
+ self.redis_instance.set(
+ "dataset_structure", json.dumps(self.datasets))
return True
return None
@@ -169,7 +172,8 @@ Publish or ProbeSet. E.g.
# This has side-effects, with the end result being a truth-y value
if(self.set_dataset_key(t, name)):
break
- return self.datasets.get(name, None) # Return None if name has not been set
+ # Return None if name has not been set
+ return self.datasets.get(name, None)
# Do the intensive work at startup one time only
@@ -651,6 +655,43 @@ class DataSet(object):
"Dataset {} is not yet available in GeneNetwork.".format(self.name))
pass
+ def fetch_probe_trait_data(self, sample_list=None):
+ if sample_list:
+ self.samplelist = sample_list
+ else:
+ self.samplelist = self.group.samplelist
+
+ if self.group.parlist != None and self.group.f1list != None:
+ if (self.group.parlist + self.group.f1list) in self.samplelist:
+ self.samplelist += self.group.parlist + self.group.f1list
+
+ query = """
+ SELECT Strain.Name, Strain.Id FROM Strain, Species
+ WHERE Strain.Name IN {}
+ and Strain.SpeciesId=Species.Id
+ and Species.name = '{}'
+ """.format(create_in_clause(self.samplelist), *mescape(self.group.species))
+ logger.sql(query)
+ results = dict(g.db.execute(query).fetchall())
+ sample_ids = [results[item] for item in self.samplelist]
+
+ query = """SELECT * from ProbeSetData WHERE Id in ( SELECT ProbeSetXRef.DataId FROM (ProbeSet, ProbeSetXRef, ProbeSetFreeze) WHERE ProbeSetXRef.ProbeSetFreezeId = ProbeSetFreeze.Id and ProbeSetFreeze.Name = 'HC_M2_0606_P' and ProbeSet.Id = ProbeSetXRef.ProbeSetId order by ProbeSet.Id ) and StrainId in ({})""".format(
+ ",".join(str(sample_id) for sample_id in sample_ids))
+
+ results = g.db.execute(query).fetchall()
+
+ # with conn:
+ # cursor = conn.cursor()
+ # cursor.execute(query)
+ # results = cursor.fetchall()
+ trait_data = {}
+ for trait_id, StrainId, value in results:
+ if trait_id in trait_data:
+ trait_data[trait_id].append(value)
+ else:
+ trait_data[trait_id] = [value]
+ self.trait_data = trait_data
+
def get_trait_data(self, sample_list=None):
if sample_list:
self.samplelist = sample_list
@@ -670,6 +711,7 @@ class DataSet(object):
logger.sql(query)
results = dict(g.db.execute(query).fetchall())
sample_ids = [results[item] for item in self.samplelist]
+ print("the number of sample ids are", len(sample_ids))
# MySQL limits the number of tables that can be used in a join to 61,
# so we break the sample ids into smaller chunks
@@ -720,6 +762,7 @@ class DataSet(object):
trait_sample_data.append(results)
trait_count = len(trait_sample_data[0])
+ print("the trait count is >>>", trait_count)
self.trait_data = collections.defaultdict(list)
# put all of the separate data together into a dictionary where the keys are