def ListPrimitives(self, request, context):
'''
List all primitives known to TA2, their IDs, versions, names, and digests. Using this
information a TA3 should know which primitives may be put into a pipeline template.
To narrow down potential primitives to use a TA3 can also ask a TA2 to do a solution
search and then observe which primitives the TA2 is using. If more metadata about primitives
is needed, then a TA3 can use the results of this call to map primitives to metadata
(from Python code or primitive annotations) on its own.
'''
list_primitives = []
source_primitives = []
primitives = index.search()
for prim in primitives:
try:
p = index.get_primitive(prim)
source_primitives.append(p)
except:
0
for p in source_primitives:
meta = p.metadata.to_json_structure()
list_primitives.append(
primitive_pb2.Primitive(id=meta['id'],
version=meta['version'],
python_path=meta['python_path'],
name=meta['name'],
digest=meta['digest']))
return core_pb2.ListPrimitivesResponse(primitives=list_primitives)
def test_entrypoint(self):
working_set_entries = copy.copy(pkg_resources.working_set.entries)
working_set_entry_keys = copy.copy(
pkg_resources.working_set.entry_keys)
working_set_by_key = copy.copy(pkg_resources.working_set.by_key)
try:
distribution = pkg_resources.Distribution(__file__)
entry_point = pkg_resources.EntryPoint.parse(
'foo2.bar2.FooBar2Primitive = test_index:FooBar2Primitive',
dist=distribution)
distribution._ep_map = {
'd3m.primitives': {
'foo2.bar2.FooBar2Primitive': entry_point
}
}
pkg_resources.working_set.add(distribution)
python_path = 'd3m.primitives.foo2.bar2.FooBar2Primitive'
self.assertIn(python_path, index.search())
self.assertIs(index.get_primitive(python_path), FooBar2Primitive)
finally:
pkg_resources.working_set.entries = working_set_entries
pkg_resources.working_set.entry_keys = working_set_entry_keys
pkg_resources.working_set.by_key = working_set_by_key
def available_primitives():
primitives_info = []
with d3m_utils.silence():
for primitive_path in d3m_index.search():
if primitive_path in PrimitivesList.BlockList:
continue
try:
primitive = d3m_index.get_primitive(primitive_path)
primitive_id = primitive.metadata.query()['id']
version = primitive.metadata.query()['version']
python_path = primitive.metadata.query()['python_path']
name = primitive.metadata.query()['name']
digest = primitive.metadata.query().get('digest', None)
primitive_info = {
'id': primitive_id,
'version': version,
'python_path': python_path,
'name': name,
'digest': digest
}
primitives_info.append(primitive_info)
except:
continue
return primitives_info
def load_from_d3m_index(self):
'''Load primitive description from installed python packages'''
for primitive_path, primitive_type in index.search().items():
primitive = self._create_primitive_desc(primitive_type.metadata)
if primitive.cls in self.black_list_package:
print('Black listing primitive: {}'.format(primitive.name))
else:
self.primitives.append(primitive)
self._setup()
def main():
ray.init(webui_host='127.0.0.1')
# Creating parser
parser = argparse.ArgumentParser(
description="Starts server from command line")
configure_parser(parser)
arguments = parser.parse_args()
# Setup logger
verbose_format = '%(asctime)s %(levelname)-8s %(processName)-15s [%(filename)s:%(lineno)d] %(message)s'
concise_format = '%(asctime)s %(levelname)-8s %(message)s'
log_format = verbose_format if arguments.verbose else concise_format
logging.basicConfig(format=log_format,
handlers=[
logging.StreamHandler(),
logging.FileHandler(
'{}/d3m.log'.format(Path.TEMP_STORAGE_ROOT),
'w', 'utf-8')
],
datefmt='%m/%d %H:%M:%S')
root_logger = logging.getLogger()
root_logger.setLevel(logging.INFO)
warnings.filterwarnings('ignore')
server = Server(arguments)
try:
load_time = time.time()
server.start()
with d3m_utils.silence():
d3m_index.load_all(blocklist=PrimitivesList.BlockList)
print('Wait for loading workers for', len(d3m_index.search()) * 0.3)
time.sleep(len(d3m_index.search()) * 0.3)
# time.sleep(5)
logger.info('---------- Waiting for Requests ----------')
while True:
time.sleep(_ONE_DAY_IN_SECONDS)
except KeyboardInterrupt:
logger.info('############ STOPPING SERVICE ############')
server.stop()
def __init__(self):
self.primitive = d3m_index.search()
self.argmentsmapper = {
"container": metadata_base.ArgumentType.CONTAINER,
"data": metadata_base.ArgumentType.DATA,
"value": metadata_base.ArgumentType.VALUE,
"primitive": metadata_base.ArgumentType.PRIMITIVE
}
self.stepcheck = None # Generate a step check matrix
self.step_number = {}
self.addstep_mapper = {
("<class 'd3m.container.pandas.DataFrame'>", "<class 'd3m.container.numpy.ndarray'>"):
"d3m.primitives.data.DataFrameToNDArray",
# ("<class 'd3m.container.pandas.DataFrame'>", "<class 'd3m.container.numpy.ndarray'>"): "d3m.primitives.data_cleaning.imputer.SKlearn",
("<class 'd3m.container.numpy.ndarray'>", "<class 'd3m.container.pandas.DataFrame'>"):
"d3m.primitives.data.NDArrayToDataFrame"
}
self.description_info = ""
self.need_add_reference = False
import os
import logging
import json
from d3m import index
logger = logging.getLogger(__name__)
PRIMITIVES_BY_NAME_PATH = os.path.join(os.path.dirname(__file__),
'../resource/primitives_by_name.json')
PRIMITIVES_BY_TYPE_PATH = os.path.join(os.path.dirname(__file__),
'../resource/primitives_by_type.json')
INSTALLED_PRIMITIVES = sorted(index.search(),
key=lambda x: x.endswith('SKlearn'),
reverse=True)
BLACK_LIST = {
'd3m.primitives.classification.random_classifier.Test',
'd3m.primitives.classification.global_causal_discovery.ClassifierRPI',
'd3m.primitives.classification.tree_augmented_naive_bayes.BayesianInfRPI',
'd3m.primitives.classification.simple_cnaps.UBC',
'd3m.primitives.classification.logistic_regression.UBC',
'd3m.primitives.classification.multilayer_perceptron.UBC',
'd3m.primitives.classification.canonical_correlation_forests.UBC',
'd3m.primitives.regression.multilayer_perceptron.UBC',
'd3m.primitives.regression.canonical_correlation_forests.UBC',
'd3m.primitives.regression.linear_regression.UBC',
'd3m.primitives.classification.inceptionV3_image_feature.Gator',
'd3m.primitives.classification.search.Find_projections',
'd3m.primitives.classification.search_hybrid.Find_projections',
'd3m.primitives.regression.search_hybrid_numeric.Find_projections',
import template_new
from d3m import utils, index
import json
primitive = index.search()
# print(primitive['d3m.primitives.datasets.Denormalize'].metadata.pretty_print())
mydsbox = template_new.DSBoxTemplate()
with open("user_defined_template_sample.json", "r") as g:
template = json.load(g)
with open("user_defined_conf_sample.json", "r") as f:
configuration_point = json.load(f)
# mypipeline = mydsbox.to_pipeline(configuration_point)
# print(configuration_point)
# print(type(configuration_point))
# print(typeof(configuration_point))
# print(mypipeline)
# template_new.printpipeline(mypipeline)
#
mytemplate = template_new.MyTemplate(template)
requirements.append("NO_CATEGORICAL_VALUES")
if ("unique" in data.name) and (not passed) and ("NOT_UNIQUE"
not in requirements):
#print("Primitive cannot handle having a column of unique values")
requirements.append("NOT_UNIQUE")
if ("negative"
in data.name) and (not passed) and ("POSITIVE_VALUES"
not in requirements):
#print("Primitive cannot handle negative values")
requirements.append("POSITIVE_VALUES")
#if(array):
# #prim.isArray = True
# prim["IsArray"] = True
prim["Requirements"] = requirements
return prim
#Main script
DATADIR = "data_profiler/" #Dir with the profiling datasets
d = {}
for primitive_name, primitive in index.search().items():
#print ("Detecting requirements for : " +primitive_name)
#if(primitive_name == "d3m.primitives.common_primitives.PCA"):
#if(primitive_name == "d3m.primitives.test.SumPrimitive"):
#print (" " + json.dumps(getPrimitiveRequirements(DATADIR,primitive_name, primitive)))
d[primitive_name] = getPrimitiveRequirements(DATADIR, primitive_name,
primitive)
print(json.dumps(d))
import pickle
from os.path import join
from d3m import index
from copy import deepcopy
from sqlalchemy.orm import joinedload
from d3m.container import Dataset
from d3m_ta2_nyu.pipeline_score import evaluate, kfold_tabular_split, score
from d3m_ta2_nyu.workflow import database
from d3m_ta2_nyu.parameter_tuning.primitive_config import is_tunable
from d3m_ta2_nyu.parameter_tuning.bayesian import HyperparameterTuning, get_new_hyperparameters
from d3m.metadata.problem import PerformanceMetric, TaskKeyword
from d3m_ta2_nyu.ta2 import create_outputfolders
logger = logging.getLogger(__name__)
PRIMITIVES = index.search()
@database.with_db
def tune(pipeline_id, metrics, problem, dataset_uri, sample_dataset_uri, report_rank, timeout_tuning, timeout_run,
msg_queue, db):
timeout_tuning = timeout_tuning * 0.9 # FIXME: Save 10% of timeout to score the best config
# Load pipeline from database
pipeline = (
db.query(database.Pipeline)
.filter(database.Pipeline.id == pipeline_id)
.options(joinedload(database.Pipeline.modules),
joinedload(database.Pipeline.connections))
).one()
logger.info('About to tune pipeline, id=%s, dataset=%r, timeout=%d secs', pipeline_id, dataset_uri, timeout_tuning)
