def test_database_from_dataframes():
with NamedTemporaryFile(suffix="test.db") as f:
db = db_from_dataframes(
db_filename=f.name,
dataframes={"A": dfA, "B": dfB},
primary_keys={"A": "numbers"},
indices={"A": [("numbers", "strings")]},
subdir="test_datacache")
cursor_A = db.execute("SELECT * FROM A")
results_A = cursor_A.fetchall()
eq_(results_A, [(1, "a"), (2, "b"), (3, "c")])
cursor_B = db.execute("SELECT * FROM B")
results_B = cursor_B.fetchall()
eq_(results_B, [("nuzzle",), ("ruzzle",)])
def test_database_from_dataframes():
with NamedTemporaryFile(suffix="test.db") as f:
db = db_from_dataframes(db_filename=f.name,
dataframes={
"A": dfA,
"B": dfB
},
primary_keys={"A": "numbers"},
indices={"A": [("numbers", "strings")]},
subdir="test_datacache")
cursor_A = db.execute("SELECT * FROM A")
results_A = cursor_A.fetchall()
eq_(results_A, [(1, "a"), (2, "b"), (3, "c")])
cursor_B = db.execute("SELECT * FROM B")
results_B = cursor_B.fetchall()
eq_(results_B, [("nuzzle", ), ("ruzzle", )])
def _create_database(self, force=False):
print("Creating database: %s" % self.local_db_path())
filename = self.local_db_filename()
df = self.gtf.dataframe()
all_index_groups = self._all_possible_indices(df.columns)
# split single DataFrame into dictionary mapping each unique
# feature name onto that subset of the data
feature_names = df["feature"].unique()
dataframes = {}
# every table gets the same set of indices
indices_dict = {}
# if a feature has an ID then make it that table's primary key
primary_keys = {}
for feature in feature_names:
df_subset = df[df.feature == feature]
dataframes[feature] = df_subset
primary_key = self._get_primary_key(feature, df_subset)
if primary_key:
primary_keys[feature] = primary_key
indices_dict[feature] = self._feature_indices(all_index_groups, primary_key, df_subset)
self._connection = datacache.db_from_dataframes(
db_filename=filename,
dataframes=dataframes,
indices=indices_dict,
primary_keys=primary_keys,
subdir=CACHE_SUBDIR,
overwrite=force,
version=DATABASE_SCHEMA_VERSION,
)
return self._connection