def test_vmerge_blind_append(self):
meta = self.example_data_A_meta
data = self.example_data_A_data
# Create left dataset
subset_columns_l = [
'unique_id', 'gender', 'locality', 'ethnicity', 'q2', 'q3'
]
meta_l, data_l = subset_dataset(meta,
data[:10],
columns=subset_columns_l)
dataset_left = (meta_l, data_l)
# Create right dataset
subset_columns_r = [
'unique_id', 'gender', 'religion', 'q1', 'q2', 'q8', 'q9'
]
meta_r, data_r = subset_dataset(meta,
data[5:15],
columns=subset_columns_r)
dataset_right = (meta_r, data_r)
# vmerge datasets indicating row_id
dataset_left = (meta_l, data_l)
meta_vm, data_vm = vmerge(dataset_left, dataset_right, verbose=False)
# check merged dataframe
verify_vmerge_data(self,
data_l,
data_r,
data_vm,
meta_vm,
blind_append=True)
def test_vmerge_basic(self):
meta = self.example_data_A_meta
data = self.example_data_A_data
# Create left dataset
subset_columns_l = [
'unique_id', 'gender', 'locality', 'ethnicity', 'q2', 'q3'
]
meta_l, data_l = subset_dataset(meta,
data[:10],
columns=subset_columns_l)
dataset_left = (meta_l, data_l)
# Create right dataset
subset_columns_r = [
'unique_id', 'gender', 'religion', 'q1', 'q2', 'q8', 'q9'
]
meta_r, data_r = subset_dataset(meta,
data[5:15],
columns=subset_columns_r)
dataset_right = (meta_r, data_r)
# vmerge datasets using left_on/right_on
dataset_left = (meta_l, data_l)
meta_vm, data_vm = vmerge(dataset_left,
dataset_right,
left_on='unique_id',
right_on='unique_id',
verbose=False)
# check merged dataframe
verify_vmerge_data(self, data_l, data_r, data_vm, meta_vm)
# vmerge datasets using on
dataset_left = (meta_l, data_l)
meta_vm, data_vm = vmerge(dataset_left,
dataset_right,
on='unique_id',
verbose=False)
# check merged data
verify_vmerge_data(self, data_l, data_r, data_vm, meta_vm)
def test_vmerge_row_id(self):
meta = self.example_data_A_meta
data = self.example_data_A_data
# Create left dataset
subset_columns_l = [
'unique_id', 'gender', 'locality', 'ethnicity', 'q2', 'q3'
]
meta_l, data_l = subset_dataset(meta,
data[:10],
columns=subset_columns_l)
dataset_left = (meta_l, data_l)
# Create right dataset
subset_columns_r = [
'unique_id', 'gender', 'religion', 'q1', 'q2', 'q8', 'q9'
]
meta_r, data_r = subset_dataset(meta,
data[5:15],
columns=subset_columns_r)
dataset_right = (meta_r, data_r)
# vmerge datasets indicating row_id
dataset_left = (meta_l, data_l)
meta_vm, data_vm = vmerge(dataset_left,
dataset_right,
on='unique_id',
row_id_name='DataSource',
left_id=1,
right_id=2,
verbose=False)
expected = {
'text': {
'en-GB': 'vmerge row id'
},
'type': 'int',
'name': 'DataSource'
}
actual = meta_vm['columns']['DataSource']
self.assertEqual(actual, expected)
self.assertTrue(data_vm['DataSource'].dtype == 'int64')
# check merged dataframe
verify_vmerge_data(self,
data_l,
data_r,
data_vm,
meta_vm,
row_id_name='DataSource',
left_id=1,
right_id=2)
# vmerge datasets indicating row_id
dataset_left = (meta_l, data_l)
meta_vm, data_vm = vmerge(dataset_left,
dataset_right,
on='unique_id',
row_id_name='DataSource',
left_id=1,
right_id=2.0,
verbose=False)
expected = {
'text': {
'en-GB': 'vmerge row id'
},
'type': 'float',
'name': 'DataSource'
}
actual = meta_vm['columns']['DataSource']
self.assertEqual(actual, expected)
self.assertTrue(data_vm['DataSource'].dtype == 'float64')
# check merged dataframe
verify_vmerge_data(self,
data_l,
data_r,
data_vm,
meta_vm,
row_id_name='DataSource',
left_id=1,
right_id=2.0)
# vmerge datasets indicating row_id
dataset_left = (meta_l, data_l)
meta_vm, data_vm = vmerge(dataset_left,
dataset_right,
on='unique_id',
row_id_name='DataSource',
left_id='W1',
right_id=2.0,
verbose=False)
expected = {
'text': {
'en-GB': 'vmerge row id'
},
'type': 'str',
'name': 'DataSource'
}
actual = meta_vm['columns']['DataSource']
self.assertEqual(actual, expected)
#self.assertTrue(data_vm['DataSource'].dtype == 'str')
self.assertTrue(is_string_dtype(data_vm['DataSource']))
# check merged dataframe
verify_vmerge_data(self,
data_l,
data_r,
data_vm,
meta_vm,
row_id_name='DataSource',
left_id='W1',
right_id='2.0')
