wukan1986/ta_cn

**版技术指标

ta_cn's Issues

COUNT 很慢

ta_cn 版本

• 0.5.0
• 通过 pip install ta_cn -i https://mirrors.aliyun.com/pypi/simple --upgrade 安装
• from ta_cn.tdx import COUNT

测试环境

用 conda 创建虚拟环境

• CentOS Linux release 7.9.2009 (Core)
• conda 4.12.0
• Python 3.10.6
• Flask 2.2.2
• gunicorn 20.1.0

测试数据

每支股票 340 条数据

测试策略

F1=COUNT(L==LLV(L,60),10)>=1;
XH=F1;

测试方式

遍历执行上面策略大概4000次（类比四千只股）

运行结果

使用 mytt 的 count 执行了 3.326728343963623s
使用 from ta_cn.tdx import COUNT 的 count 执行了 8.872456312179565s

alpha191中alpha036 公式错误

RANK(SUM(CORR(RANK(VOLUME), RANK(VWAP)), 6), 2)中corr没有Period, 以我的理解应该改为：
RANK(SUM(CORR(RANK(VOLUME), RANK(VWAP), 6), 2))

Alpha10公式有疑问

原先这个RANK是没有参数的，源码中传入5应该被忽略
原先研报中是否是
先取5日MAX，再取RANK
(RANK(TSMAX(((RET < 0) ? STD(RET, 20) : CLOSE)^2),5)))

Alpha143公式计算有问题

def alpha_143(CLOSE, **kwargs):
"""Alpha143 CLOSE>DELAY(CLOSE,1)?(CLOSE-DELAY(CLOSE,1))/DELAY(CLOSE,1)*SELF:SELF"""
t1 = DELAY(CLOSE, 1)
t2 = IF(CLOSE > t1, CLOSE / t1 - 1., 1.)
return CUMPROD(t2)

没有看到对于self（上一期结果）的处理

0.5.0 缺少 numba bottleneck

运行环境

• centos7.9
• python 3.10.7

开启虚拟环境

python3 -m venv venv

alpha83 报错

Traceback (most recent call last):
  File "/Users/xxx/code/futudata/main.py", line 46, in <module>
    ret = func(**kwargs)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/alphas/alpha191.py", line 622, in alpha_083
    t1 = COVIANCE(RANK(HIGH), RANK(VOLUME), 5)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/utils_long.py", line 88, in decorated
    s3 = s2.groupby(by=by, group_keys=False).apply(to_pd(dataframe_split(func)), **_kwargs)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/groupby/groupby.py", line 1423, in apply
    result = self._python_apply_general(f, self._selected_obj)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/groupby/groupby.py", line 1464, in _python_apply_general
    values, mutated = self.grouper.apply(f, data, self.axis)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/groupby/ops.py", line 761, in apply
    res = f(group)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/groupby/groupby.py", line 1397, in f
    return func(g, *args, **kwargs)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/utils.py", line 14, in decorated
    return np_to_pd(func(*args, **kwargs),
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/utils_long.py", line 107, in decorated
    return func(*args_input, *args, **kwargs_input, **kwargs)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/statistics.py", line 79, in COVAR
    return numpy_rolling_apply([pd_to_np(real0), pd_to_np(real1)],
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/nb.py", line 289, in numpy_rolling_apply
    return func1(*arrs, out, window, func2, *args)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/numba/core/dispatcher.py", line 468, in _compile_for_args
    error_rewrite(e, 'typing')
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/numba/core/dispatcher.py", line 409, in error_rewrite
    raise e.with_traceback(None)
numba.core.errors.TypingError: Failed in nopython mode pipeline (step: nopython frontend)
Failed in nopython mode pipeline (step: nopython frontend)
No implementation of function Function(<function cov at 0x10408ecb0>) found for signature:
 
 >>> cov(readonly array(float64, 1d, A), readonly array(float64, 1d, A))
 
There are 2 candidate implementations:
  - Of which 2 did not match due to:
  Overload in function 'np_cov': File: numba/np/arraymath.py: Line 2888.
    With argument(s): '(readonly array(float64, 1d, A), readonly array(float64, 1d, A))':
   Rejected as the implementation raised a specific error:
     TypingError: Failed in nopython mode pipeline (step: nopython frontend)
   No implementation of function Function(<function np_cov_impl_inner at 0x148201240>) found for signature:
    
    >>> np_cov_impl_inner(array(float64, 2d, C), Literal[bool](False), none)
    
   There are 2 candidate implementations:
         - Of which 2 did not match due to:
         Overload in function 'register_jitable.<locals>.wrap.<locals>.ov_wrap': File: numba/core/extending.py: Line 159.
           With argument(s): '(array(float64, 2d, C), bool, none)':
          Rejected as the implementation raised a specific error:
            TypingError: Failed in nopython mode pipeline (step: nopython frontend)
          No implementation of function Function(<function dot at 0x103259870>) found for signature:
           
           >>> dot(array(float64, 2d, C), array(float64, 2d, F))
           
          There are 4 candidate implementations:
                - Of which 4 did not match due to:
                Overload in function '_OverloadWrapper._build.<locals>.ol_generated': File: numba/core/overload_glue.py: Line 131.
                  With argument(s): '(array(float64, 2d, C), array(float64, 2d, F))':
                 Rejected as the implementation raised a specific error:
                   LoweringError: Failed in nopython mode pipeline (step: native lowering)
                 scipy 0.16+ is required for linear algebra
                 
                 File "<string>", line 3:
                 <source missing, REPL/exec in use?>
                 
                 During: lowering "$8call_function.3 = call $2load_global.0(tmp0, tmp1, func=$2load_global.0, args=[Var(tmp0, <string>:3), Var(tmp1, <string>:3)], kws=(), vararg=None, target=None)" at <string> (3)
            raised from /Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/numba/core/errors.py:837
          
          During: resolving callee type: Function(<function dot at 0x103259870>)
          During: typing of call at /Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/numba/np/arraymath.py (2739)
          
          
          File "../../.conda/envs/futudata/lib/python3.10/site-packages/numba/np/arraymath.py", line 2739:
          def np_cov_impl_inner(X, bias, ddof):
              <source elided>
              # calculate result - requires blas
              c = np.dot(X, np.conj(X.T))
              ^
   
     raised from /Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/numba/core/typeinfer.py:1086
   
   During: resolving callee type: Function(<function np_cov_impl_inner at 0x148201240>)
   During: typing of call at /Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/numba/np/arraymath.py (2927)
   
   
   File "../../.conda/envs/futudata/lib/python3.10/site-packages/numba/np/arraymath.py", line 2927:
       def np_cov_impl(m, y=None, rowvar=True, bias=False, ddof=None):
           <source elided>
           else:
               return np_cov_impl_inner(X, bias, ddof)
               ^

  raised from /Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/numba/core/typeinfer.py:1086

During: resolving callee type: Function(<function cov at 0x10408ecb0>)
During: typing of call at /Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/nb.py (207)


File "../../.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/nb.py", line 207:
def _cov_nb(arr0, arr1):
    <source elided>
    """协方差矩阵"""
    return _np.cov(arr0, arr1)[0, 1]
    ^

During: resolving callee type: type(CPUDispatcher(<function _cov_nb at 0x13709edd0>))
During: typing of call at /Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/nb.py (239)

During: resolving callee type: type(CPUDispatcher(<function _cov_nb at 0x13709edd0>))
During: typing of call at /Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/nb.py (239)


File "../../.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/nb.py", line 239:
def _rolling_func_2_nb(arr0, arr1, out, timeperiod, func, *args):
    <source elided>
        for i, (a, b) in enumerate(zip(arr0, arr1)):
            out[i + timeperiod - 1] = func(a, b, *args)
            ^

python3 跑 alpha075 报错

Traceback (most recent call last):
  File "/Users/xxx/code/futudata/main.py", line 57, in <module>
    ret = l.alpha_075(**kwargs_l)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/ta_cn/alphas/test191.py", line 505, in alpha_075
    return COUNT((CLOSE > OPEN) & (BANCHMARKINDEXCLOSE < BANCHMARKINDEXOPEN), 50) / COUNT(
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/ops/common.py", line 70, in new_method
    return method(self, other)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/arraylike.py", line 124, in __truediv__
    return self._arith_method(other, operator.truediv)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/series.py", line 5639, in _arith_method
    return base.IndexOpsMixin._arith_method(self, other, op)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/base.py", line 1295, in _arith_method
    result = ops.arithmetic_op(lvalues, rvalues, op)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/ops/array_ops.py", line 220, in arithmetic_op
    _bool_arith_check(op, left, right)
  File "/Users/xxx/.conda/envs/futudata/lib/python3.10/site-packages/pandas/core/ops/array_ops.py", line 520, in _bool_arith_check
    raise NotImplementedError(
NotImplementedError: operator 'truediv' not implemented for bool dtypes

单只股票的alpha191没办法求

报一下错误
File "D:\06_projects\python\stock_trade\alphas191.py", line 62, in <module> alpha_001 = l.alpha_001(OPEN=OPEN, CLOSE=CLOSE, VOLUME=VOLUME) File "D:\ProgramData\Anaconda3\envs\py38\lib\site-packages\ta_cn\alphas\alpha191.py", line 14, in alpha_001 return (-1 * CORR(RANK(DELTA(LOG(VOLUME), 1)), RANK(((CLOSE - OPEN) / OPEN)), 6)) File "D:\ProgramData\Anaconda3\envs\py38\lib\site-packages\ta_cn\utils_long.py", line 112, in decorated return pd.Series(s3, index=s1.index) File "D:\ProgramData\Anaconda3\envs\py38\lib\site-packages\pandas\core\series.py", line 386, in __init__ if is_empty_data(data) and dtype is None: File "D:\ProgramData\Anaconda3\envs\py38\lib\site-packages\pandas\core\construction.py", line 877, in is_empty_data is_simple_empty = is_list_like_without_dtype and not data File "D:\ProgramData\Anaconda3\envs\py38\lib\site-packages\pandas\core\generic.py", line 1527, in __nonzero__ raise ValueError( ValueError: The truth value of a DataFrame is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

而且，dataframe的索引还必须设置asset，不然报错，这是什么原因？难道不能单独求值？