def many_materializations_and_passing_expectations(_context):
tables = [
'users',
'groups',
'events',
'friends',
'pages',
'fans',
'event_admins',
'group_admins',
]
for table in tables:
yield Materialization(
label='table_info',
metadata_entries=[
EventMetadataEntry.path(label='table_path',
path='/path/to/{}.raw'.format(table))
],
)
yield ExpectationResult(
success=True,
label='{table}.row_count'.format(table=table),
description='Row count passed for {table}'.format(table=table),
)
def backcompat_materialize(_):
yield Materialization(
asset_key="all_types",
description="a materialization with all metadata types",
metadata_entries=[
MetadataEntry("text", value="text is cool"),
MetadataEntry(
"url", value=MetadataValue.url("https://bigty.pe/neato")),
MetadataEntry("path",
value=MetadataValue.path("/tmp/awesome")),
MetadataEntry("json", value={"is_dope": True}),
MetadataEntry(
"python class",
value=MetadataValue.python_artifact(MetadataEntry)),
MetadataEntry(
"python function",
value=MetadataValue.python_artifact(file_relative_path)),
MetadataEntry("float", value=1.2),
MetadataEntry("int", value=1),
MetadataEntry("float NaN", value=float("nan")),
MetadataEntry("long int", value=LONG_INT),
MetadataEntry("pipeline run",
value=MetadataValue.pipeline_run("fake_run_id")),
MetadataEntry("my asset", value=AssetKey("my_asset")),
],
)
yield Output(None)
def sort_by_calories(context, cereals):
sorted_cereals = sorted(
cereals, key=lambda cereal: int(cereal['calories'])
)
context.log.info(
'Least caloric cereal: {least_caloric}'.format(
least_caloric=sorted_cereals[0]['name']
)
)
context.log.info(
'Most caloric cereal: {most_caloric}'.format(
most_caloric=sorted_cereals[-1]['name']
)
)
fieldnames = list(sorted_cereals[0].keys())
sorted_cereals_csv_path = os.path.abspath(
'output/calories_sorted_{run_id}.csv'.format(run_id=context.run_id)
)
os.makedirs(os.path.dirname(sorted_cereals_csv_path), exist_ok=True)
with open(sorted_cereals_csv_path, 'w') as fd:
writer = csv.DictWriter(fd, fieldnames)
writer.writeheader()
writer.writerows(sorted_cereals)
yield Materialization(
label='sorted_cereals_csv',
description='Cereals data frame sorted by caloric content',
metadata_entries=[
EventMetadataEntry.path(
sorted_cereals_csv_path, 'sorted_cereals_csv_path'
)
],
)
yield Output(None)
def train_lstm_model(context, training_set: TrainingSet):
X, y = training_set
breakpoint = context.solid_config['timeseries_train_test_breakpoint'] # pylint: disable=W0622
X_train, X_test = X[0:breakpoint], X[breakpoint:]
y_train, y_test = y[0:breakpoint], y[breakpoint:]
_, n_steps, n_features = X.shape
model = Sequential()
model.add(
LSTM(
context.solid_config['lstm_layer_config']['num_recurrant_units'],
activation=context.solid_config['lstm_layer_config']['activation'],
input_shape=(n_steps, n_features),
))
model.add(Dense(context.solid_config['num_dense_layers']))
model.compile(
optimizer=context.solid_config['model_trainig_config']['optimizer'],
loss=context.solid_config['model_trainig_config']['loss'],
metrics=['mae'],
)
model.fit(
X_train,
y_train,
epochs=context.solid_config['model_trainig_config']['num_epochs'],
verbose=0,
)
results = model.evaluate(X_test, y_test, verbose=0)
yield Materialization(
label='test_set_results',
metadata_entries=[
EventMetadataEntry.text(str(results[0]), 'Mean Squared Error'),
EventMetadataEntry.text(str(results[1]), 'Mean Absolute Error'),
],
)
yield Output(model)
def write_operation_inventory(context: SolidExecutionContext,
analysis: Dict[str, RightSizeAnalysis],
resources: DataFrame) -> Nothing:
resources = resources.set_index('resource_id')
resizes = [{
'subscription_id': resources.at[resource_id, 'subscription_id'],
'resource_id': resource_id,
'current_sku': resources.at[resource_id, 'vm_size'],
'new_sku': analysis.advisor_sku
} for resource_id, analysis in analysis.items()
if analysis.advisor_sku_valid]
output = {'vm_resize_operations': resizes}
output_path = os.path.abspath(f'operation_inventory_{context.run_id}.json')
with open(output_path, 'w') as fd:
json.dump(output, fd, indent=3)
yield Materialization(
label='operation_inventory',
description=
'An inventory of the right sizing operations that are recommended and validated.',
metadata_entries=[
EventMetadataEntry.path(output_path, 'operation_inventory_path')
],
)
yield Output(None)
def logs_events(context):
context.log_event(AssetMaterialization("first"))
context.log_event(Materialization("second"))
context.log_event(ExpectationResult(success=True))
context.log_event(AssetObservation("fourth"))
yield AssetMaterialization("fifth")
yield Output("blah")
def course_roles(context: SolidExecutionContext,
edx_course_ids: List[String]) -> DagsterPath:
"""Retrieve information about user roles for given courses.
:param context: Dagster execution context for propagaint configuration data
:type context: SolidExecutionContext
:param edx_course_ids: List of edX course ID strings
:type edx_course_ids: List[String]
:returns: A path definition that points to the rendered data table
:rtype: DagsterPath
"""
access_role = Table('student_courseaccessrole')
roles_query = Query.from_(access_role).select(
'id', 'user_id', 'org', 'course_id',
'role').where(access_role.course_id.isin(edx_course_ids))
query_fields, roles_data = context.resources.sqldb.run_query(roles_query)
# Maintaining previous file name for compatibility (TMM 2020-05-01)
roles_path = context.resources.results_dir.path.joinpath('role_query.csv')
write_csv(query_fields, roles_data, roles_path)
yield Materialization(
label='role_query.csv',
description='Course roles records from Open edX installation',
metadata_entries=[
EventMetadataEntry.text(
label='course_roles_count',
description='Number of course roles records',
text=str(len(roles_data))),
EventMetadataEntry.path(roles_path.name, 'role_query_csv_path')
])
yield Output(roles_path, 'edx_course_roles')
def df_output_schema(_context, path, value):
with open(path, 'w') as fd:
writer = csv.DictWriter(fd, fieldnames=value[0].keys())
writer.writeheader()
writer.writerows(rowdicts=value)
return Materialization.file(path)
def spark_df_output_schema(_context, file_type, file_options, spark_df):
if file_type == 'csv':
spark_df.write.csv(file_options['path'],
header=file_options.get('header'),
sep=file_options.get('sep'))
return Materialization.file(file_options['path'])
else:
check.failed('Unsupported file type: {}'.format(file_type))
def save_to_file_materialization(_, cfg, value):
with open(cfg['path'], 'w') as ff:
ff.write(str(value))
return Materialization(
'path',
'Wrote out value to {path}'.format(path=path),
metadata_entries=[EventMetadataEntry.text('path', path)],
)
def spark_df_materializer(_context, config, spark_df):
file_type, file_options = list(config.items())[0]
if file_type == 'csv':
spark_df.write.csv(**file_options)
return Materialization.file(file_options['path'])
elif file_type == 'parquet':
spark_df.write.parquet(**file_options)
return Materialization.file(file_options['path'])
elif file_type == 'json':
spark_df.write.json(**file_options)
return Materialization.file(file_options['path'])
elif file_type == 'jdbc':
spark_df.write.jdbc(**file_options)
return Materialization.file(file_options['url'])
elif file_type == 'orc':
spark_df.write.orc(**file_options)
return Materialization.file(file_options['path'])
elif file_type == 'saveAsTable':
spark_df.write.saveAsTable(**file_options)
return Materialization.file(file_options['name'])
elif file_type == 'text':
spark_df.write.text(**file_options)
return Materialization.file(file_options['path'])
else:
check.failed('Unsupported file type: {}'.format(file_type))
def raw_file_solid(_context):
yield Materialization(
label='table_info',
metadata_entries=[
EventMetadataEntry.path(label='table_path', path='/path/to/{}.raw'.format(name))
],
)
yield do_expectation(_context, name)
yield Output(name)
def load_data_to_database_from_spark(context, data_frame):
context.resources.db_info.load_table(data_frame,
context.solid_config['table_name'])
# TODO Flow more information down to the client
# We should be able to flow multiple key value pairs down to dagit
# See https://github.com/dagster-io/dagster/issues/1408
yield Materialization(path='Persisted Db Table: {table_name}'.format(
table_name=context.solid_config['table_name']))
yield Result(data_frame)
def less_simple_data_frame_output_materialization_config(
context, config, value):
# Materialize LessSimpleDataFrame into a csv file
csv_path = os.path.join(os.path.dirname(__file__),
os.path.abspath(config['csv']['path']))
os.makedirs(os.path.dirname(csv_path), exist_ok=True)
with open(csv_path, 'w') as fd:
fieldnames = list(value[0].keys())
writer = csv.DictWriter(fd, fieldnames, delimiter=config['csv']['sep'])
writer.writeheader()
writer.writerows(value)
context.log.debug(
'Wrote dataframe as .csv to {path}'.format(path=csv_path))
yield Materialization(
'1data_frame_csv',
'LessSimpleDataFrame materialized as csv',
[
EventMetadataEntry.path(
path=csv_path,
label='data_frame_csv_path',
description='LessSimpleDataFrame written to csv format',
)
],
)
# Materialize LessSimpleDataFrame into a json file
json_path = os.path.abspath(config['json']['path'])
with open(json_path, 'w') as fd:
json_value = seven.json.dumps([dict(row) for row in value])
fd.write(json_value)
context.log.debug(
'Wrote dataframe as .json to {path}'.format(path=json_path))
yield Materialization(
'data_frame_json',
'LessSimpleDataFrame materialized as json',
[
EventMetadataEntry.path(
path=json_path,
label='data_frame_json_path',
description='LessSimpleDataFrame written to json format',
)
],
)
def insert_into_staging_table(context, records: DataFrame, table_name: str):
_create_and_load_staging_table(context.resources.postgres_db.engine, table_name, records)
yield Materialization(
label=table_name,
description='Table {} created in database {}'.format(
table_name, context.resources.postgres_db.db_name
),
metadata_entries=[EventMetadataEntry.text(str(len(records)), "num rows inserted")],
)
yield Output(output_name='staging_table', value=table_name)
def train_lstm_model_and_upload_to_gcs(context, training_set: TrainingSet,
bucket_name: str):
from keras.layers import LSTM, Dense
from keras.models import Sequential
X, y = training_set
breakpoint = context.solid_config['timeseries_train_test_breakpoint'] # pylint: disable=W0622
X_train, X_test = X[0:breakpoint], X[breakpoint:]
y_train, y_test = y[0:breakpoint], y[breakpoint:]
_, n_steps, n_features = X.shape
model = Sequential()
model.add(
LSTM(
context.solid_config['lstm_layer_config']['num_recurrant_units'],
activation=context.solid_config['lstm_layer_config']['activation'],
input_shape=(n_steps, n_features),
))
model.add(Dense(context.solid_config['num_dense_layers']))
model.compile(
optimizer=context.solid_config['model_trainig_config']['optimizer'],
loss=context.solid_config['model_trainig_config']['loss'],
metrics=['mae'],
)
model.fit(
X_train,
y_train,
epochs=context.solid_config['model_trainig_config']['num_epochs'],
verbose=0,
)
results = model.evaluate(X_test, y_test, verbose=0)
# save model and upload
gcs_bucket = context.resources.gcs_client.get_bucket(bucket_name)
key = 'model-{}.h5'.format(uuid.uuid4())
with tempfile.TemporaryFile('w+b') as fp:
model.save(fp)
# Done because you can't upload the contents of a file outside the context manager if it's a tempfile.
fp.seek(0)
gcs_bucket.blob(key).upload_from_file(fp)
yield Materialization(
description='Serialized model to Google Cloud Storage Bucket',
label='Serialized model and uploaded to gcs',
metadata_entries=[
EventMetadataEntry.text(
'gs://{bucket_name}/{key}'.format(bucket_name=bucket_name,
key=key),
'google cloud storage URI',
),
EventMetadataEntry.text(str(results[0]), 'Mean Squared Error'),
EventMetadataEntry.text(str(results[1]), 'Mean Absolute Error'),
],
)
yield Output(model)
def emit_events_solid(_, input_num):
a_num = input_num + 1
a_string = 'foo'
yield ExpectationResult(
success=a_num > 0, label='positive', description='A num must be positive'
)
yield Materialization(
label='persisted_string', description='Let us pretend we persisted the string somewhere'
)
yield Output(value=a_num, output_name='a_num')
yield Output(value=a_string, output_name='a_string')
def materialize_one(_):
yield Materialization(
label='one',
asset_key=asset_key,
metadata_entries=[
EventMetadataEntry.text('hello', 'text'),
EventMetadataEntry.json({'hello': 'world'}, 'json'),
EventMetadataEntry.float(1.0, 'one'),
],
)
yield Output(1)
def materialize(_):
yield Materialization(
label='all_types',
description='a materialization with all metadata types',
metadata_entries=[
EventMetadataEntry.text('text is cool', 'text'),
EventMetadataEntry.url('https://bigty.pe/neato', 'url'),
EventMetadataEntry.fspath('/tmp/awesome', 'path'),
EventMetadataEntry.json({'is_dope': True}, 'json'),
],
)
yield Output(None)
def _base_compute(context):
time.sleep(context.solid_config['sleep'])
if random() < context.solid_config['error_rate']:
raise Exception('blah')
if context.solid_config.get('materialization_key') is not None:
metadata_entries = []
if context.solid_config.get('materialization_text') is not None:
metadata_entries.append(
EventMetadataEntry.text(
context.solid_config.get('materialization_text'), context.solid.name,
)
)
if context.solid_config.get('materialization_url') is not None:
metadata_entries.append(
EventMetadataEntry.url(
context.solid_config.get('materialization_url'), context.solid.name,
)
)
if context.solid_config.get('materialization_path') is not None:
metadata_entries.append(
EventMetadataEntry.path(
context.solid_config.get('materialization_url'), context.solid.name,
)
)
if context.solid_config.get('materialization_json') is not None:
metadata_entries.append(
EventMetadataEntry.json(
context.solid_config.get('materialization_json'), context.solid.name,
)
)
if context.solid_config.get('materialization_value') is not None:
metadata_entries = [
EventMetadataEntry.float(
context.solid_config.get('materialization_value'), context.solid.name,
)
]
if len(metadata_entries) == 0:
metadata_entries = None
yield Materialization(
label=context.solid.name,
asset_key=context.solid_config.get('materialization_key'),
metadata_entries=metadata_entries,
)
yield Output(1)
def my_asset_key_materialization_solid(context, df):
do_some_transform(df)
persist_to_storage(df)
yield Materialization(
asset_key=AssetKey(['dashboard', 'my_cool_site']),
description='Persisted result to storage',
metadata_entries=[
EventMetadataEntry.url('http://mycoolsite.com/dashboard',
label='dashboard_url'),
EventMetadataEntry.float(calculate_bytes(df), 'size (bytes)'),
],
)
yield Output(df)
def send_to_slack(context, download_data):
transaction_data = download_data[0]
block_data = download_data[1]
transaction_date = transaction_data['transaction_date'][0]
block_date = block_data['block_date'][0]
yield ExpectationResult(
label='dates_match',
success=transaction_date == block_date,
metadata_entries=[
EventMetadataEntry.text(str(transaction_date), 'transaction_date'),
EventMetadataEntry.text(str(block_date), 'block_date'),
],
)
date = transaction_date
dash_transferred = transaction_data['DASH_transferred'][0]
dash_blocks = block_data['DASH_blocks'][0]
average_dash_transferred_per_block = float(dash_transferred) / dash_blocks
yield Materialization(
label='data',
metadata_entries=[
EventMetadataEntry.text(
'{dash_transferred} dash tranferred'.format(
dash_transferred=dash_transferred),
'dash_transferred',
),
EventMetadataEntry.text(
'{dash_blocks} dash blocks'.format(dash_blocks=dash_blocks),
'dash_blocks'),
],
)
context.resources.slack.chat.post_message(
channel='#metrics-testing',
text=
'{date}\nDash Transferred: {dash_transferred}\nDash blocks: {dash_blocks}\n'
'Average dash transferred/block: {average_dash_transferred_per_block}'.
format(
date=date,
dash_transferred=dash_transferred,
dash_blocks=dash_blocks,
average_dash_transferred_per_block=
average_dash_transferred_per_block,
),
)
yield Output(1)
def try_parse_create_view(text):
view_match = CREATE_VIEW_REGEX.search(text)
if not view_match:
return None
return Materialization(
label='create_view',
description=text,
metadata_entries=[
EventMetadataEntry.text(view_match.group(1), 'schema'),
EventMetadataEntry.text(view_match.group(2), 'view'),
],
)
def many_table_materializations(_context):
for table in raw_tables:
yield Materialization(
label='table_info',
metadata_entries=[
EventMetadataEntry.text(text=table, label='table_name'),
EventMetadataEntry.fspath(path='/path/to/{}'.format(table),
label='table_path'),
EventMetadataEntry.json(data={'name': table},
label='table_data'),
EventMetadataEntry.url(url='https://bigty.pe/{}'.format(table),
label='table_name_big'),
],
)
def materialization_and_expectation(_context):
yield Materialization(
label='all_types',
description='a materialization with all metadata types',
metadata_entries=[
EventMetadataEntry.text('text is cool', 'text'),
EventMetadataEntry.url('https://bigty.pe/neato', 'url'),
EventMetadataEntry.fspath('/tmp/awesome', 'path'),
EventMetadataEntry.json({'is_dope': True}, 'json'),
],
)
yield ExpectationResult(success=True, label='row_count', description='passed')
yield ExpectationResult(True)
yield Output(True)
def write_html_report(context: SolidExecutionContext,
report_notebook: FileHandle) -> Nothing:
with context.file_manager.read(report_notebook) as node_file:
node = nbformat.read(node_file, nbformat.NO_CONVERT)
html = convert_nodebook_node_to_html(node, full_width=True)
handle = context.file_manager.write_data(html.encode(), ext='html')
yield Materialization(
label='resize_report',
description=
'A report of all VMs utilization data and evaulation of the recommendations.',
metadata_entries=[
EventMetadataEntry.path(handle.path_desc, 'resize_report_path')
],
)
def file_handle_to_s3(context, file_handle):
bucket = context.solid_config['Bucket']
key = context.solid_config['Key']
with context.file_manager.read(file_handle, 'rb') as fileobj:
context.resources.s3.upload_fileobj(fileobj, bucket, key)
s3_file_handle = S3FileHandle(bucket, key)
yield Materialization(
label='file_to_s3',
metadata_entries=[EventMetadataEntry.path(s3_file_handle.s3_path, label=last_key(key))],
)
yield Output(value=s3_file_handle, output_name='s3_file_handle')
def try_parse_create_table(text):
table_match = CREATE_TABLE_REGEX.search(text)
if not table_match:
return None
return Materialization(
label='create_table',
description=text,
metadata_entries=[
EventMetadataEntry.text(table_match.group(1), 'schema'),
EventMetadataEntry.text(table_match.group(2), 'table'),
EventMetadataEntry.text(table_match.group(3), 'row_count'),
],
)
def less_simple_data_frame_output_materialization_config(
context, config, value):
csv_path = os.path.abspath(config['csv']['path'])
with open(csv_path, 'w') as fd:
fieldnames = list(value[0].keys())
writer = csv.DictWriter(fd, fieldnames, delimiter=config['csv']['sep'])
writer.writeheader()
writer.writerows(value)
context.log.debug(
'Wrote dataframe as .csv to {path}'.format(path=csv_path))
return Materialization(
'data_frame_csv',
'LessSimpleDataFrame materialized as csv',
[EventMetadataEntry.path(csv_path, 'data_frame_csv_path')],
)
def load_data_to_database_from_spark(context, data_frame: DataFrame):
context.resources.db_info.load_table(data_frame, context.solid_config['table_name'])
table_name = context.solid_config['table_name']
yield Materialization(
label='Table: {table_name}'.format(table_name=table_name),
description=(
'Persisted table {table_name} in database configured in the db_info resource.'
).format(table_name=table_name),
metadata_entries=[
EventMetadataEntry.text(label='Host', text=context.resources.db_info.host),
EventMetadataEntry.text(label='Db', text=context.resources.db_info.db_name),
],
)
yield Output(value=table_name, output_name='table_name')
