n = 3000
m = 3000
index = pd.date_range(dt.datetime(1990, 1, 1),
dt.datetime(1990, 1, 1) + dt.timedelta(days=m - 1))
index = np.repeat(index, n)
df = pd.DataFrame(np.random.randn(n * m, 3),
columns=['x', 'y', 'z'],
index=index)
df['c'] = matlib.repmat(np.linspace(0, n - 1, n, dtype=int), 1, m)[0]
start = dt.datetime.now()
t = MA(20, 'x') / MA(30, 'y')
res = t.transform(df, category_field='c')
print("Finance-Python (analysis): {0}s".format(dt.datetime.now() - start))
start = dt.datetime.now()
groups = df.groupby('c')
res = groups['x'].rolling(20).mean() / groups['y'].rolling(30).mean()
print("Pandas (group by): {0}s".format(dt.datetime.now() - start))
start = dt.datetime.now()
t = MovingAverage(20, 'x') / MovingAverage(30, 'x')
res = t.transform(df)
print("Finance-Python (accumulator): {0}s".format(dt.datetime.now() - start))
start = dt.datetime.now()
res = df['x'].rolling(20).mean() / df['x'].rolling(30).mean()
print("Pandas (group by): {0}s".format(dt.datetime.now() - start))
# -*- coding: utf-8 -*-
u"""
Created on 2016-12-25
@author: cheng.li
"""
import datetime as dt
import pandas as pd
sample_data = pd.DataFrame(
data={'code': [1, 2, 1, 2, 1, 2],
'open': [2.0, 1.0, 1.5, 3.0, 2.4, 3.5],
'close': [1.7, 1.6, 0.9, 3.8, 1.6, 2.1]},
index=[dt.datetime(2016, 1, 1),
dt.datetime(2016, 1, 1),
dt.datetime(2016, 1, 2),
dt.datetime(2016, 1, 2),
dt.datetime(2016, 1, 3),
dt.datetime(2016, 1, 3)]
)
sample_data = sample_data[['code', 'open', 'close']]
if __name__ == '__main__':
from PyFin.api import MA
ts = MA(2, 'close')
res = ts.transform(sample_data, name='ma_2_no_code')
print(res)