# -*- coding: utf-8 -*-
from pandas import DataFrame
from pandas_ta.overlap import rma
from pandas_ta.utils import get_offset, non_zero_range, verify_series
[docs]def kdj(high=None, low=None, close=None, length=None, signal=None, offset=None, **kwargs):
"""Indicator: KDJ (KDJ)"""
# Validate Arguments
length = int(length) if length and length > 0 else 9
signal = int(signal) if signal and signal > 0 else 3
_length = max(length, signal)
high = verify_series(high, _length)
low = verify_series(low, _length)
close = verify_series(close, _length)
offset = get_offset(offset)
if high is None or low is None or close is None: return
# Calculate Result
highest_high = high.rolling(length).max()
lowest_low = low.rolling(length).min()
fastk = 100 * (close - lowest_low) / non_zero_range(highest_high, lowest_low)
k = rma(fastk, length=signal)
d = rma(k, length=signal)
j = 3 * k - 2 * d
# Offset
if offset != 0:
k = k.shift(offset)
d = d.shift(offset)
j = j.shift(offset)
# Handle fills
if "fillna" in kwargs:
k.fillna(kwargs["fillna"], inplace=True)
d.fillna(kwargs["fillna"], inplace=True)
j.fillna(kwargs["fillna"], inplace=True)
if "fill_method" in kwargs:
k.fillna(method=kwargs["fill_method"], inplace=True)
d.fillna(method=kwargs["fill_method"], inplace=True)
j.fillna(method=kwargs["fill_method"], inplace=True)
# Name and Categorize it
_params = f"_{length}_{signal}"
k.name = f"K{_params}"
d.name = f"D{_params}"
j.name = f"J{_params}"
k.category = d.category = j.category = "momentum"
# Prepare DataFrame to return
kdjdf = DataFrame({k.name: k, d.name: d, j.name: j})
kdjdf.name = f"KDJ{_params}"
kdjdf.category = "momentum"
return kdjdf
kdj.__doc__ = \
"""KDJ (KDJ)
The KDJ indicator is actually a derived form of the Slow
Stochastic with the only difference being an extra line
called the J line. The J line represents the divergence
of the %D value from the %K. The value of J can go
beyond [0, 100] for %K and %D lines on the chart.
Sources:
https://www.prorealcode.com/prorealtime-indicators/kdj/
https://docs.anychart.com/Stock_Charts/Technical_Indicators/Mathematical_Description#kdj
Calculation:
Default Inputs:
length=9, signal=3
LL = low for last 9 periods
HH = high for last 9 periods
FAST_K = 100 * (close - LL) / (HH - LL)
K = RMA(FAST_K, signal)
D = RMA(K, signal)
J = 3K - 2D
Args:
high (pd.Series): Series of 'high's
low (pd.Series): Series of 'low's
close (pd.Series): Series of 'close's
length (int): Default: 9
signal (int): Default: 3
offset (int): How many periods to offset the result. Default: 0
Kwargs:
fillna (value, optional): pd.DataFrame.fillna(value)
fill_method (value, optional): Type of fill method
Returns:
pd.Series: New feature generated.
"""