def test_join(self):
# Nothing too technical, do not need to test pandas functions
set_string(self.url_pval, self.valid_workflow_URL)
set_string(self.colnames_pval, 'key')
set_integer(self.type_pval, _join_type_map.index("inner"))
result = JoinURL.render(self.wfm, self.table)
assert_frame_equal(self.ref_left_join, result.dataframe)
def test_type_mismatch(self):
# Joining on column with different dtypes should fail
set_string(self.url_pval, self.valid_workflow_URL)
set_string(self.colnames_pval, 'key')
set_integer(self.type_pval, _join_type_map.index("inner"))
result = JoinURL.render(self.wfm, self.table_with_types)
self.assertEqual(("Types do not match for key column 'key' (number and text). " \
'Please use a type conversion module to make these column types consistent.'), result.error)
def test_type_num_cast(self):
# Joining on column that is both int and float, module should convert to float
version = self.wfm.store_fetched_table(self.ext_workflow_with_types)
self.wfm.set_fetched_data_version(version)
set_string(self.url_pval, self.valid_workflow_URL)
set_string(self.colnames_pval, 'key')
set_integer(self.type_pval, _join_type_map.index("inner"))
result = JoinURL.render(self.wfm, self.table_with_types)
assert_frame_equal(self.ref_left_join_with_types, result.dataframe)
def test_invalid_url(self):
version = self.wfm.store_fetched_table(self.csv_table)
self.wfm.set_fetched_data_version(version)
set_string(self.url_pval, 'not a url')
result = ConcatURL.render(self.wfm, self.table)
self.assertTrue('URL' in result.error)
# Working, non-workflow URL should fail
set_string(self.url_pval, 'www.google.com')
result = ConcatURL.render(self.wfm, self.table)
self.assertTrue('URL' in result.error)
def test_concat_with_source(self):
version = self.wfm.store_fetched_table(self.ext_workflow)
self.wfm.set_fetched_data_version(version)
set_string(self.url_pval, self.valid_workflow_URL)
set_checkbox(self.source_columns_pval, True)
set_integer(self.type_pval, _type_map.index("include columns from both workflows"))
result = ConcatURL.render(self.wfm, self.table)
# Sanitize dataframe to clean index created by pandas concat()
result.sanitize_in_place()
ref = self.ref_outer_concat.copy()
ref.insert(0, _source_column_name, ['Current', 'Current', '2', '2'])
assert_frame_equal(ref, result.dataframe)
def test_concat(self):
version = self.wfm.store_fetched_table(self.ext_workflow)
self.wfm.set_fetched_data_version(version)
set_string(self.url_pval, self.valid_workflow_URL)
set_checkbox(self.source_columns_pval, False)
set_integer(self.type_pval, _type_map.index("only include this workflow's columns"))
result = ConcatURL.render(self.wfm, self.table)
# Sanitize dataframe to clean index created by pandas concat()
result.sanitize_in_place()
assert_frame_equal(self.ref_source_only_concat, result.dataframe)
set_integer(self.type_pval, _type_map.index("only include matching columns"))
result = ConcatURL.render(self.wfm, self.table)
# Sanitize dataframe to clean index created by pandas concat()
result.sanitize_in_place()
assert_frame_equal(self.ref_inner_concat, result.dataframe)
set_integer(self.type_pval, _type_map.index("include columns from both workflows"))
result = ConcatURL.render(self.wfm, self.table)
# Sanitize dataframe to clean index created by pandas concat()
result.sanitize_in_place()
assert_frame_equal(self.ref_outer_concat, result.dataframe)
def test_columns_no_exist(self):
# Should throw error for select column not existing
set_checkbox(self.select_columns_pval, True)
set_string(self.importcols_pval, 'non_existent_col')
set_string(self.colnames_pval, 'key')
result = JoinURL.render(self.wfm, self.table)
self.assertEqual(
"Import columns not in target workflow: {'non_existent_col'}",
result.error)
# Should throw error for select column not existing
set_string(self.importcols_pval, '')
set_string(self.colnames_pval, 'non_existent_col')
result = JoinURL.render(self.wfm, self.table)
self.assertEqual(
"Key columns not in target workflow: {'non_existent_col'}",
result.error)
def test_importcols(self):
# Should only import 1 column
set_string(self.url_pval, self.valid_workflow_URL)
set_string(self.colnames_pval, 'key')
set_string(self.importcols_pval, 'col2')
set_checkbox(self.select_columns_pval, True)
set_integer(self.type_pval, _join_type_map.index("inner"))
result = JoinURL.render(self.wfm, self.table.copy())
assert_frame_equal(self.ref_left_join[['col1', 'key', 'col2']],
result.dataframe)
# When select_column false, should import all columns
set_checkbox(self.select_columns_pval, False)
result = JoinURL.render(self.wfm, self.table.copy())
assert_frame_equal(self.ref_left_join, result.dataframe)
def test_tsv(self):
set_string(self.csv_pval, count_tsv)
result = execute_wfmodule(self.wf_module)
self.assertEqual(result, ProcessResult(reference_table))
def test_empty(self):
set_string(self.csv_pval, '')
result = execute_wfmodule(self.wf_module)
self.assertEqual(result, ProcessResult(pd.DataFrame()))