def pinbar_detector(data: pd.DataFrame,
labelBuy: str = "PIN_Buy",
labelSell: str = "PIN_Sell",
labelOpen: str = "<OPEN>",
labelHigh: str = "<HIGH>",
labelLow: str = "<LOW>",
labelClose: str = "<CLOSE>") -> pd.DataFrame:
"""
Pinbar detector takes dataframe with price data (OHLC) and produces new columns
with information if given record is a pinbar signal for buy or sell.
Args:
data (pd.DataFrame): input dataframe with price data
labelBuy (str, optional): label used for pin buy signal column. Defaults to "PIN_Buy".
labelSell (str, optional): label used for pin sell signal column. Defaults to "PIN_Sell".
labelOpen (str, optional): Open price label in input dataframe. Defaults to "<OPEN>".
labelHigh (str, optional): High price label in input dataframe. Defaults to "<HIGH>".
labelLow (str, optional): Low price label in input dataframe. Defaults to "<LOW>".
labelClose (str, optional): Close pirce label in input dataframe. Defaults to "<CLOSE>".
Returns:
pd.DataFrame: Input dataframe with added columns for detected signals.
"""
data[labelBuy] = np.zeros((len(data.index), 1), dtype=int)
data[labelSell] = np.zeros((len(data.index), 1), dtype=int)
data, tempColumns = calculate_body_and_tails(data, labelOpen, labelHigh,
labelLow, labelClose)
if (tempColumns):
data.loc[((data["tailDown"] > 3 * data["body"].abs()) &
(data["tailUp"] < 0.5 * data["tailDown"])), labelBuy] = 1
data.loc[((data["tailUp"] > 3 * data["body"].abs()) &
(data["tailDown"] < 0.5 * data["tailUp"])), labelSell] = 1
data = data.drop(tempColumns, axis=1, errors='ignore')
return data
def test_if_tailUp_is_calculated(dataframe_with_test_data):
dataframe_out = calculate_body_and_tails(dataframe_with_test_data)[0]
dataframe_out.loc[
dataframe_out["body"] > 0,
"tailUpGT"] = dataframe_out["<HIGH>"] - dataframe_out["<CLOSE>"]
dataframe_out.loc[
dataframe_out["body"] <= 0,
"tailUpGT"] = dataframe_out["<HIGH>"] - dataframe_out["<OPEN>"]
assert dataframe_out["tailUp"].equals(dataframe_out["tailUpGT"])
def fakey_detector(data: pd.DataFrame,
labelBuy: str = "Fakey_Buy",
labelSell: str = "Fakey_Sell",
labelOpen: str = "<OPEN>",
labelHigh: str = "<HIGH>",
labelLow: str = "<LOW>",
labelClose: str = "<CLOSE>") -> pd.DataFrame:
"""
Fakey detector takes dataframe with price data (OHLC) and produces new columns
with information if given record is a fakey signal for buy or sell.
Args:
data (pd.DataFrame): input dataframe with price data
labelBuy (str, optional): label used for fakey buy signal column. Defaults to "Fakey_Buy".
labelSell (str, optional): label used for fakey sell signal column. Defaults to "Fakey_Sell".
labelOpen (str, optional): Open price label in input dataframe. Defaults to "<OPEN>".
labelHigh (str, optional): High price label in input dataframe. Defaults to "<HIGH>".
labelLow (str, optional): Low price label in input dataframe. Defaults to "<LOW>".
labelClose (str, optional): Close pirce label in input dataframe. Defaults to "<CLOSE>".
Returns:
pd.DataFrame: Input dataframe with added columns for detected signals.
"""
data[labelBuy] = np.zeros((len(data.index), 1), dtype=int)
data[labelSell] = np.zeros((len(data.index), 1), dtype=int)
data, tempColumns = calculate_body_and_tails(data, labelOpen, labelHigh,
labelLow, labelClose)
if (tempColumns):
data["body-1"] = data["body"].shift(1, fill_value=0)
data["close-1"] = data[labelClose].shift(1, fill_value=0)
data["open-1"] = data[labelOpen].shift(1, fill_value=0)
tempColumns += ["body-1", "close-1", "open-1"]
data.loc[(
(data["body-1"] > 0)
& (data[labelHigh] > data["close-1"] + data["body-1"].abs() * 0.25)
& (data[labelOpen] < data["close-1"])
& (data[labelClose] < data["close-1"])
& (data[labelOpen] > data["open-1"])
& (data[labelClose] > data["open-1"])), labelSell] = 1
data.loc[(
(data["body-1"] < 0)
& (data[labelLow] < data["close-1"] - data["body-1"].abs() * 0.25)
& (data[labelOpen] > data["close-1"])
& (data[labelClose] > data["close-1"])
& (data[labelOpen] < data["open-1"])
& (data[labelClose] < data["open-1"])), labelBuy] = 1
data = data.drop(tempColumns, axis=1, errors='ignore')
return data
def test_if_tails_columns_are_added():
dataframe_out = calculate_body_and_tails(
pd.DataFrame([], columns=testColumns))[0]
assert "tailUp" in dataframe_out.columns
assert "tailDown" in dataframe_out.columns
def test_if_added_columns_are_in_list():
dataframe_out, added_columns = calculate_body_and_tails(
pd.DataFrame([], columns=testColumns))
for name in dataframe_out.columns:
if name not in testColumns:
assert name in added_columns
def test_if_body_is_calculated(dataframe_with_test_data):
dataframe_out = calculate_body_and_tails(dataframe_with_test_data)[0]
dataframe_out[
"bodyGT"] = dataframe_out["<CLOSE>"] - dataframe_out["<OPEN>"]
assert dataframe_out["body"].equals(dataframe_out["bodyGT"])
def test_if_calculate_body_and_tails_returns_dataframe_with_column_body_added(
):
dataframe_out = calculate_body_and_tails(
pd.DataFrame([], columns=testColumns))[0]
assert "body" in dataframe_out.columns
def test_if_calculate_body_and_tails_inputs_and_outputs_dataframe():
dataframe_out = pd.DataFrame([])
assert type(calculate_body_and_tails(dataframe_out)[0]) is pd.DataFrame
assert type(calculate_body_and_tails(dataframe_out)[1]) is list