# -*- coding: utf-8 -*-
from numpy import isnan
from pandas import DataFrame, Series
from pandas_ta._typing import DictLike, Int, IntFloat
from pandas_ta.ma import ma
from pandas_ta.overlap import ema
from pandas_ta.utils import (
    v_drift,
    v_mamode,
    v_offset,
    v_pos_default,
    v_scalar,
    v_series
)



def tsi(
    close: Series, fast: Int = None, slow: Int = None,
    signal: Int = None, scalar: IntFloat = None,
    mamode: str = None, drift: Int = None,
    offset: Int = None, **kwargs: DictLike
) -> DataFrame:
    """True Strength Index

    This indicator attempts to identify short-term swings in trend direction
    as well as identifying possible "overbought" and "oversold" signals.

    Sources:
        * [investopedia](https://www.investopedia.com/terms/t/tsi.asp)

    Parameters:
        close (Series): ```close``` Series
        fast (int): Fast MA period. Default: ```13```
        slow (int): Slow MA period. Default: ```25```
        signal (int): Signal period. Default: ```13```
        scalar (float): Scalar. Default: ```100```
        mamode (str): Signal MA. See ```help(ta.ma)```. Default: ```"ema"```
        drift (int): Difference amount. Default: ```1```
        offset (int): Post shift. Default: ```0```

    Other Parameters:
        fillna (value): ```pd.DataFrame.fillna(value)```

    Returns:
        (DataFrame): 2 columns
    """
    # Validate
    fast = v_pos_default(fast, 13)
    slow = v_pos_default(slow, 25)
    signal = v_pos_default(signal, 13)
    if slow < fast:
        fast, slow = slow, fast
    _length = slow + signal + 1
    close = v_series(close, _length)

    if "length" in kwargs:
        kwargs.pop("length")

    if close is None:
        return

    scalar = v_scalar(scalar, 100)
    mamode = v_mamode(mamode, "ema")
    drift = v_drift(drift)
    offset = v_offset(offset)

    # Calculate
    diff = close.diff(drift)
    slow_ema = ema(close=diff, length=slow, **kwargs)
    if all(isnan(slow_ema)):
        return  # Emergency Break
    fast_slow_ema = ema(close=slow_ema, length=fast, **kwargs)

    abs_diff = diff.abs()
    abs_slow_ema = ema(close=abs_diff, length=slow, **kwargs)
    if all(isnan(abs_slow_ema)):
        return  # Emergency Break
    abs_fast_slow_ema = ema(close=abs_slow_ema, length=fast, **kwargs)

    tsi = scalar * fast_slow_ema / abs_fast_slow_ema
    if all(isnan(tsi)):
        return  # Emergency Break
    tsi_signal = ma(mamode, tsi, length=signal)

    # Offset
    if offset != 0:
        tsi = tsi.shift(offset)
        tsi_signal = tsi_signal.shift(offset)

    # Fill
    if "fillna" in kwargs:
        tsi.fillna(kwargs["fillna"], inplace=True)
        tsi_signal.fillna(kwargs["fillna"], inplace=True)

    # Name and Category
    tsi.name = f"TSI_{fast}_{slow}_{signal}"
    tsi_signal.name = f"TSIs_{fast}_{slow}_{signal}"
    tsi.category = tsi_signal.category = "momentum"

    data = {tsi.name: tsi, tsi_signal.name: tsi_signal}
    df = DataFrame(data, index=close.index)
    df.name = f"TSI_{fast}_{slow}_{signal}"
    df.category = "momentum"

    return df