def test_is_close_up(self):
df = pdDF(data={
'open': [1., 2., 3., 4.],
'close': [1.1, 2.1, 0.1, 4.1]
})
val = KlineFeatures.is_close_up(df)
assert np.allclose(val, [True, True, False, True])
def test_ochl_features(self):
df = pdDF(
data={
'open': [1., 2., 3., 4.],
'close': [1.1, 2.1, 3.1, 4.1],
'high': [1.2, 2.2, 3.3, 4.4],
'low': [0.1, 0.2, 0.3, 0.4]
})
o, c, h, l = KlineFeatures.ochl_features(df)
assert o[0] == 1.
assert c[0] == 1.1
assert h[0] == 1.2
assert l[0] == 0.1
def update_realtime_stock_info(self) -> pdDF:
stock_list = []
for k, v in self._monitored_stock_map.items():
stock_list.extend(v)
if not stock_list:
return pdDF()
try:
self._df_realtime_stock_info = sina_api.get_realtime_stock_infos(
stock_list)
except:
mylog.warn(f'Can not query realtime info from sina api')
return self._df_realtime_stock_info
def extract_result(content_list: str):
find_results = re.finditer(r'var hq_str_(..\d{6})="(.*)";',
content_list)
stock_codes = []
attrs = []
for v in find_results:
stock_codes.append(CodeTools.to_pcode(v[1]))
content = v[2].split(',')
content = content[0:6]
content.append(content.pop(0))
attrs.append(content)
df = pdDF(attrs, index=stock_codes, columns=_column_names, dtype=float)
df['name'] = df['name'].astype(str)
return df
def test_day_compare_features(self):
df = pdDF(
data={
'open': [1., 2., 3., 4.],
'close': [1.1, 2.1, 3.1, 4.1],
'high': [1., 2.1, 3.2, 4.3],
'low': [0.1, 0.2, 0.3, 0.4]
})
fs = KlineFeatures.day_compare_features(df)
result = np.asarray([*fs])
test_result = np.asarray([[
1.,
0.95,
0.93333333,
0.925,
], [
1.1,
1.,
0.96666667,
0.95,
], [0., 0.45, 0.6, 0.675], [
0.,
0.5,
0.63333333,
0.7,
], [1., 1., 1., 1.], [
0.1,
0.05,
0.03333333,
0.025,
], [
0.,
0.45,
0.63333333,
0.725,
], [0., 0., 0., 0.]])
assert np.all(result - test_result < 0.0001)
def _np_compound_arr_to_df(arr):
val = pdDF(arr)
val['datetime'] = (val['datetime'] / 1000000).astype(numpy.int64)
val.set_index('datetime', inplace=True)
return val
def test_up_day_count_of_previous_days(self):
df = pdDF(data={'close': np.asarray([0, 1, 2, 1, 0, 1, 0])})
val = KlineFeatures.up_count_of_previous_days(df)
np.array_equal(val, [0, 0, 1, 2, 0, 0, 1])
def test_jump_value(self):
df = pdDF(data={'high': [2, 4, 1.3, 1.4], 'low': [0.1, 2.2, 0.3, 0.4]})
jump_values = KlineFeatures.jump_value(df)
assert np.allclose(jump_values, [0.1, 0.2, -0.9, 0.])
def test_is_up_inday(self):
df = pdDF(data={'open': [1, 2, 0], 'close': [2, 1, 0]})
val = KlineFeatures.is_up_inday(df)
np.array_equal(val, [True, False, False])
