def test_market_data_object():
coord_val_pair = [
{
'coordinate': {
'mkt_type': 'IR',
'mkt_asset': 'USD',
'mkt_class': 'Swap',
'mkt_point': ('5y', ),
'mkt_quoting_style': 'ATMRate'
},
'value': 0.9973194889
},
{
'coordinate': {
'mkt_type': 'IR',
'mkt_asset': 'USD',
'mkt_class': 'Swap',
'mkt_point': ('40y', ),
'mkt_quoting_style': 'ATMRate'
},
'value': 'redacted'
},
]
coordinates = {
MarketDataCoordinate.from_dict(dic['coordinate']): dic['value']
for dic in coord_val_pair
}
overlay_market = OverlayMarket(base_market=CloseMarket(
market_data=coordinates))
assert overlay_market.coordinates[0] == MarketDataCoordinate.from_dict(
coord_val_pair[0]['coordinate'])
assert overlay_market.redacted_coordinates[
0] == MarketDataCoordinate.from_dict(coord_val_pair[1]['coordinate'])
def handle_result(result: Optional[Union[DataFrameWithInfo, ErrorValue, PricingFuture]]) -> \
[OverlayMarket, dict]:
if isinstance(result, ErrorValue):
return result
properties = MarketDataCoordinate.properties()
is_historical = result.index.name == 'date'
location = PricingContext.current.market_data_location
def extract_market_data(this_result: DataFrameWithInfo):
market_data = {}
for _, row in this_result.iterrows():
coordinate_values = {p: row.get(p) for p in properties}
mkt_point = coordinate_values.get('mkt_point')
if mkt_point is not None:
coordinate_values['mkt_point'] = tuple(coordinate_values['mkt_point'].split(';'))
# return 'redacted' as coordinate value if its a redacted coordinate
market_data[MarketDataCoordinate.from_dict(coordinate_values)] = row['value'] if \
row['permissions'] == 'Granted' else 'redacted'
return market_data
if is_historical:
return {date: OverlayMarket(
base_market=CloseMarket(date=date, location=location),
market_data=extract_market_data(result.loc[date]))
for date in set(result.index)}
else:
return OverlayMarket(base_market=result.risk_key.market,
market_data=extract_market_data(result))
def handle_result(result: Optional[Union[DataFrameWithInfo, ErrorValue, PricingFuture]]) ->\
[OverlayMarket, dict]:
properties = MarketDataCoordinate.properties()
is_historical = isinstance(PricingContext.current,
HistoricalPricingContext)
location = PricingContext.current.market_data_location
def extract_market_data(this_result: DataFrameWithInfo):
market_data = {}
for _, row in result.iterrows():
coordinate_values = {p: row.get(p) for p in properties}
if 'mkt_point' in coordinate_values:
coordinate_values['mkt_point'] = tuple(
coordinate_values['mkt_point'].split(';'))
market_data[MarketDataCoordinate.from_dict(
coordinate_values)] = row['value']
return market_data
if is_historical:
return {
date: OverlayMarket(
base_market=CloseMarket(date=date, location=location),
market_data=extract_market_data(result.loc[date]))
for date in set(result.index)
}
else:
return OverlayMarket(base_market=result.risk_key.market,
market_data=extract_market_data(result))
def extract_market_data(this_result: DataFrameWithInfo):
market_data = {}
for _, row in result.iterrows():
coordinate_values = {p: row.get(p) for p in properties}
if 'mkt_point' in coordinate_values:
coordinate_values['mkt_point'] = tuple(
coordinate_values['mkt_point'].split(';'))
market_data[MarketDataCoordinate.from_dict(
coordinate_values)] = row['value']
return market_data
def extract_market_data(this_result: DataFrameWithInfo):
market_data = {}
for _, row in this_result.iterrows():
coordinate_values = {p: row.get(p) for p in properties}
mkt_point = coordinate_values.get('mkt_point')
if mkt_point is not None:
coordinate_values['mkt_point'] = tuple(coordinate_values['mkt_point'].split(';'))
# return 'redacted' as coordinate value if its a redacted coordinate
market_data[MarketDataCoordinate.from_dict(coordinate_values)] = row['value'] if \
row['permissions'] == 'Granted' else 'redacted'
return market_data
def _coordinate_from_str(cls, coordinate_str: str) -> MarketDataCoordinate:
tmp = coordinate_str.rsplit(".", 1)
dimensions = tmp[0].split("_")
if len(dimensions) < 2:
raise MqValueError('invalid coordinate ' + coordinate_str)
kwargs = {
'mkt_type': dimensions[0],
'mkt_asset': dimensions[1] or None,
'mkt_quoting_style': tmp[-1] if len(tmp) > 1 else None}
if len(dimensions) > 2:
kwargs['mkt_class'] = dimensions[2] or None
if len(dimensions) > 3:
kwargs['mkt_point'] = tuple(dimensions[3:]) or None
return MarketDataCoordinate(**kwargs)
def get_many_coordinates(
cls,
mkt_type: str = None,
mkt_asset: str = None,
mkt_class: str = None,
mkt_point: Tuple[str, ...] = (),
*,
limit: int = 100,
return_type: type = str,
) -> Union[Tuple[str, ...], Tuple[MarketDataCoordinate, ...]]:
where = FieldFilterMap(
mkt_type=mkt_type.upper() if mkt_type is not None else None,
mkt_asset=mkt_asset.upper() if mkt_asset is not None else None,
mkt_class=mkt_class.upper() if mkt_class is not None else None,
)
for index, point in enumerate(mkt_point):
setattr(where, 'mkt_point' + str(index + 1), point.upper())
query = EntityQuery(
where=where,
limit=limit
)
results = GsSession.current._post('/data/mdapi/query', query)['results']
if return_type is str:
return tuple(coordinate['name'] for coordinate in results)
elif return_type is MarketDataCoordinate:
return tuple(
MarketDataCoordinate(
mkt_type=coordinate['dimensions']['mktType'],
mkt_asset=coordinate['dimensions']['mktAsset'],
mkt_class=coordinate['dimensions']['mktClass'],
mkt_point=tuple(coordinate['dimensions']['mktPoint'].values()),
mkt_quoting_style=coordinate['dimensions']['mktQuotingStyle']
) for coordinate in results)
else:
raise NotImplementedError('Unsupported return type')
import datetime as dt
import pandas as pd
import pytest
from pandas.util.testing import assert_frame_equal, assert_series_equal
from gs_quant.api.gs.data import GsDataApi
from gs_quant.context_base import ContextMeta
from gs_quant.errors import MqValueError
from gs_quant.markets import MarketDataCoordinate
from gs_quant.session import GsSession, Environment
from gs_quant.target.data import MDAPIDataQuery, MarketDataVendor, DataSetEntity
test_coordinates = (
MarketDataCoordinate(mkt_type='Prime',
mkt_quoting_style='price',
mkt_asset='335320934'),
MarketDataCoordinate(mkt_type='IR',
mkt_asset='USD',
mkt_class='Swap',
mkt_point=('2Y', )),
)
test_str_coordinates = ('Prime_335320934_.price', 'IR_USD_Swap_2Y')
test_defn_dict = {
'id': 'EXAMPLE_FROM_SLANG',
'name': 'Example DataSet',
'description': 'This is a test.',
'shortDescription': '',
'vendor': 'Goldman Sachs',
"""
import pytest
from pandas.util.testing import assert_frame_equal
from gs_quant.api.gs.data import GsDataApi
from gs_quant.markets import MarketDataCoordinate
from gs_quant.risk import sort_risk
import datetime as dt
import pandas as pd
from gs_quant.session import GsSession, Environment
from gs_quant.context_base import ContextMeta
test_coordinates = (
MarketDataCoordinate('Prime', field='price', marketDataAsset='335320934'),
MarketDataCoordinate('IR',
marketDataAsset='USD',
pointClass='Swap',
marketDataPoint=('2Y', ),
quotingStyle='ATMRate'),
)
def test_coordinates_data(mocker):
bond_data = [{
'marketDataType': 'Prime',
'marketDataAsset': '335320934',
'field': 'price',
'price': 1.0139,
'time': pd.to_datetime('2019-01-20T01:03:00Z')