def market_data_panel(self, freq="D"):
start, end = self.date_range
df = DataAPI.candle(self.universe, start=start, end=end,
freq=freq).transpose(2, 1, 0)
if freq == "D":
df.major_axis = df.major_axis.normalize()
return df
def read_benchmark(start, end, index_code="000300.XSHG", freq="D"):
"""
获取指数行情数据
:param start:起始时间(datetime)
:param end:结束时间(datetime)
:param index_code:指数代码(str)
:param freq:数据频率(str)
:return:指数行情。返回一个benchmark对象
包含五个属性。"index", "close", "open", "high", "low"。
"index":一个MultiIndex Series的序列。索引为date (level 0) 和 asset (level 1),包含一列factor值。其中asset的值固定是index_code,
factor值固定是1。用于fxdayu_alphaman.selector.selector_analysis.get_stocklist_mean_return中
作为stock_list参数来计算指数收益。
"close":收盘价。(pandas.Dateframe ),index为datetime,column.name 为index_code,值为对应指数的收盘价。
"open":开盘价。(pandas.Dateframe ),index为datetime,column.name 为index_code,值为对应指数的开盘价。
"high":最高价。(pandas.Dateframe ),index为datetime,column.name 为index_code,值为对应指数的最高价。
"low":最低价。(pandas.Dateframe ),index为datetime,column.name 为index_code,值为对应指数的最低价。
"""
benchmark = Benchmark()
benchmark_value = DataAPI.candle((index_code,), freq=freq, start=start, end=end)
benchmark.open = benchmark_value.minor_xs("open")
if len(benchmark.open)==0:
raise ValueError
benchmark.high = benchmark_value.minor_xs("high")
benchmark.low = benchmark_value.minor_xs("low")
benchmark.close = benchmark_value.minor_xs("close")
benchmark.open.index.name = "date"
benchmark.high.index.name = "date"
benchmark.low.index.name = "date"
benchmark.close.index.name = "date"
benchmark.index = pd.DataFrame(data=benchmark.open.index)
benchmark.index["asset"] = index_code
benchmark.index["factor"] = 1
benchmark.index = benchmark.index.set_index(["date", "asset"])
return benchmark
def get_price_data(pool, start, end, max_window=10):
data = DataAPI.candle(
tuple(pool),
"D",
start=start - datetime.timedelta(days=max_window),
end=end + datetime.timedelta(days=max_window))
data = data.replace(to_replace=0, value=np.NaN)
return data
def read_LFLO(symbols, start, end):
"""
获取对数流通市值
:param symbols: 一组股票代码(list),形式为通用标准(编码.交易所 如["000001.XSHE","600000.XSHG"])
:param start: 开始时间(datetime)
:param end:结束时间(datetime)
:return: 对数流通市值。(pandas.Dataframe) index为datetime,columns为股票代码,值为对应股票对应时间的对数流通市值。
"""
LFLO = DataAPI.factor(tuple(symbols), fields="LFLO", start=start,
end=end).minor_xs("LFLO")
return LFLO
def market_data(self):
start, end = self.date_range
df = DataAPI.candle(self.universe,
fields=["close"],
start=start,
end=end,
freq=self._freq).loc[:, :, "close"].T
df.index.name = "order_book_id"
series = df.stack()
series.name = "close"
return series
def benchmark_rets(self, code="000300.XSHG", freq="D", isIndex=True):
start, end = self.date_range
adjust = None
if not isIndex:
adjust = "before"
benchmark = DataAPI.candle(code,
freq=freq,
fields="close",
start=start,
end=end,
adjust=adjust)
benchmark_rets = benchmark["close"].pct_change()
benchmark_rets.name = "benchmark_rets"
if freq == "D":
benchmark_rets.index = benchmark_rets.index.normalize()
return benchmark_rets
def execute(self, pool, start, end, data=None, data_config=None):
"""
计算选股结果
:param pool: list 待选股票池
data：进行选股操作所需要的数据
start:datetime 选股索引范围起始时间
end：datetime 选股索引范围结束时间
:return:selector_result:A MultiIndex Series indexed by date (level 0) and asset (level 1), containing
the values mean whether choose the asset or not.
-----------------------------------
date | asset |
-----------------------------------
| AAPL | 1
-----------------------
| BA | 1
-----------------------
2014-01-01 | CMG | 1
-----------------------
| DAL | 0
-----------------------
| LULU | -1
-----------------------
"""
if data is None:
data = DataAPI.get(symbols=tuple(pool),
start=start -
datetime.timedelta(days=self.max_window),
end=end,
**data_config)
factor = Factor_Volume003().get_factor(data, update=True)
quantiles = self.get_quantiles(factor)
selector_result = quantiles.copy()
selector_result[:] = 0
selector_result[quantiles == self.choose_quantile] = 1
return selector_result.loc[start:end]
def execute(self, pool, start, end, data=None, data_config=None):
"""
计算选股结果
:param pool: list 待选股票池
data：进行选股操作所需要的数据
start:datetime 选股索引范围起始时间
end：datetime 选股索引范围结束时间
:return:selector_result:A MultiIndex Series indexed by date (level 0) and asset (level 1), containing
the values mean whether choose the asset or not.
-----------------------------------
date | asset |
-----------------------------------
| AAPL | 1
-----------------------
| BA | 1
-----------------------
2014-01-01 | CMG | 1
-----------------------
| DAL | 0
-----------------------
| LULU | -1
-----------------------
"""
if data is None:
data = DataAPI.get(symbols=tuple(pool),
start=start -
datetime.timedelta(days=self.max_window),
end=end,
**data_config)
selector_result = list(
map(self.calculate_MACD_signal, data.iteritems()))
selector_result = pd.concat(selector_result, axis=1).stack()
selector_result.index.names = ["date", "asset"]
return selector_result.loc[start:end]
# 配置选股器所在包路径
Admin.PACKAGE_NAME = "fxdayu_alphaman.examples.factors"
# 初始选股范围设置
initial_codes = standard_code_style(
json.load(open('test_stock_pool.json'))["test_stock_pool"])
data_config = {"freq": "D", "api": "candle", "adjust": "after"}
# 测试参数设置
start = datetime.datetime(2017, 1, 1)
end = datetime.datetime(2017, 4, 18, 15)
periods = (1, 5, 10)
# 获取数据
data = DataAPI.get(symbols=tuple(initial_codes),
start=start - datetime.timedelta(days=100),
end=end,
**data_config)
prices = data.minor_xs("close")
def manage_factors_value_test(factor_name_list, data_config_dict):
# admin测试 -获得多个因子结果
factor_admin = Admin(*factor_name_list)
factors_dict = factor_admin.get_all_factors_value(
initial_codes,
start,
end,
all_factors_data_config_dict=data_config_dict)
return factor_admin, factors_dict
def instantiate_factor_and_get_factor_value(factor_name,
pool,
start,
end,
Factor=None,
data=None,
data_config={"freq": "D", "api": "candle", "adjust": "after"},
para_dict=None,
):
"""
计算某个因子指定时间段的因子值 (可支持直接用factor_name加载到对应因子的算法)
:param factor_name: 因子名称(str) 需确保传入的factor_name、因子的类名、对应的module文件名一致(不含.后缀),因子才能正确加载
:param pool: 股票池范围(list),如：["000001.XSHE","600300.XSHG",......]
:param start: 起始时间 (datetime)
:param end: 结束时间 (datetime)
:param Factor (optional): 因子(factor.factor.Factor object),可选.可以输入一个设计好的Factor类来执行计算.
:param data (optional): 计算因子需用到的数据,根据计算需求自行指定。(可选)
:param data_config (optional): 在data参数为None的情况下(不传入自定义数据),
可通过该参数调用fxdayu_data api 访问到数据 (dict),
与data参数二选一。
:param para_dict (optional): 外部指定因子里所用到的参数集(dict),为空则不修改原有参数。 形如:{"fast":5,"slow":10}
:return: factor_value:因子值 格式为一个MultiIndex Series，索引(index)为date(level 0)和asset(level 1),
包含一列factor值。形如：
-----------------------------------
date | asset |
-----------------------------------
| AAPL | 0.5
-----------------------
| BA | -1.1
-----------------------
2014-01-01 | CMG | 1.7
-----------------------
| DAL | -0.1
-----------------------
| LULU | 2.7
-----------------------
"""
from fxdayu_data import DataAPI
import datetime
# 实例化因子类
if Factor is None:
factor = _get_factor(factor_name, Admin.PACKAGE_NAME)()
else:
factor = Factor
# 接收外部传入的参数
if para_dict:
for para in list(para_dict.keys()):
setattr(factor, para, para_dict[para])
if data is None:
pn_data = DataAPI.get(symbols=tuple(pool),
start=start - datetime.timedelta(days=factor.max_window),
end=end,
**data_config)
else:
pn_data = data
# 因子计算结果获取
factor_value = factor.get_factor(pn_data, update=True)
return factor_value
def read_benchmark(start, end, index_code="000300.XSHG", freq="D"):
"""
获取指数行情数据
:param start:起始时间(datetime)
:param end:结束时间(datetime)
:param index_code:指数代码(str)
:param freq:数据频率(str)
:return:指数行情。返回一个benchmark对象
包含五个属性。"index", "close", "open", "high", "low"。
"index":一个MultiIndex Series的序列。索引为date (level 0) 和 asset (level 1),包含一列factor值。其中asset的值固定是index_code,
factor值固定是1。用于fxdayu_alphaman.selector.selector_analysis.get_stocklist_mean_return中
作为stock_list参数来计算指数收益。
"close":收盘价。(pandas.Dateframe ),index为datetime,column.name 为index_code,值为对应指数的收盘价。
"open":开盘价。(pandas.Dateframe ),index为datetime,column.name 为index_code,值为对应指数的开盘价。
"high":最高价。(pandas.Dateframe ),index为datetime,column.name 为index_code,值为对应指数的最高价。
"low":最低价。(pandas.Dateframe ),index为datetime,column.name 为index_code,值为对应指数的最低价。
"""
benchmark = Benchmark()
try:
benchmark_value = DataAPI.candle((index_code, ),
freq=freq,
start=start,
end=end)
benchmark.open = benchmark_value.minor_xs("open")
if len(benchmark.open) == 0:
raise ValueError
benchmark.high = benchmark_value.minor_xs("high")
benchmark.low = benchmark_value.minor_xs("low")
benchmark.close = benchmark_value.minor_xs("close")
benchmark.open.index.name = "date"
benchmark.high.index.name = "date"
benchmark.low.index.name = "date"
benchmark.close.index.name = "date"
except:
index_value = ts.get_k_data(code=index_code[0:6],
start=start.strftime("%Y-%m-%d"),
end=end.strftime("%Y-%m-%d"),
ktype=freq,
index=True)
date = index_value.pop('date')
index_value["date"] = pd.to_datetime(date + " 15:00:00",
format='%Y-%m-%d %H:%M:%S')
benchmark.close = index_value[["date", "close"]]
benchmark.open = index_value[["date", "open"]]
benchmark.high = index_value[["date", "high"]]
benchmark.low = index_value[["date", "low"]]
benchmark.close.columns = ["date", index_code]
benchmark.open.columns = ["date", index_code]
benchmark.high.columns = ["date", index_code]
benchmark.low.columns = ["date", index_code]
benchmark.close = benchmark.close.set_index("date")
benchmark.open = benchmark.open.set_index("date")
benchmark.high = benchmark.high.set_index("date")
benchmark.low = benchmark.low.set_index("date")
benchmark.index = pd.DataFrame(data=benchmark.open.index)
benchmark.index["asset"] = index_code
benchmark.index["factor"] = 1
benchmark.index = benchmark.index.set_index(["date", "asset"])
return benchmark
from fxdayu_data import DataAPI
print(DataAPI.candle("000001.XSHE"))
