PVI update to match Investopedia definition

pandas_ta/core.py

@@ -1806,10 +1806,10 @@ def obv(self, offset=None, **kwargs: DictLike):
         result = obv(close=close, volume=volume, offset=offset, **kwargs)
         return self._post_process(result, **kwargs)
 
-    def pvi(self, length=None, initial=None, signed=True, offset=None, **kwargs: DictLike):
+    def pvi(self, length=None, initial=None, mamode=None, offset=None, **kwargs: DictLike):
         close = self._get_column(kwargs.pop("close", "close"))
         volume = self._get_column(kwargs.pop("volume", "volume"))
-        result = pvi(close=close, volume=volume, length=length, initial=initial, signed=signed, offset=offset, **kwargs)
+        result = pvi(close=close, volume=volume, length=length, initial=initial, mamode=mamode, offset=offset, **kwargs)
         return self._post_process(result, **kwargs)
 
     def pvo(self, fast=None, slow=None, signal=None, scalar=None, offset=None, **kwargs: DictLike):

pandas_ta/volume/pvi.py

@@ -1,14 +1,16 @@
 # -*- coding: utf-8 -*-
+import pandas as pd
+import numpy as np
 from pandas import Series
 from pandas_ta._typing import DictLike, Int
+from pandas_ta.ma import ma
 from pandas_ta.momentum import roc
-from pandas_ta.utils import signed_series, v_offset, v_pos_default, v_series
-
+from pandas_ta.utils import signed_series, v_offset, v_mamode, v_pos_default, v_series
 
 def pvi(
     close: Series, volume: Series, length: Int = None, initial: Int = None,
-    offset: Int = None, **kwargs: DictLike
-) -> Series:
+    mamode: str = None, offset: Int = None, **kwargs: DictLike
+) -> pd.DataFrame:
     """Positive Volume Index (PVI)
 
     The Positive Volume Index is a cumulative indicator that uses volume
@@ -21,48 +23,68 @@ def pvi(
     Args:
         close (pd.Series): Series of 'close's
         volume (pd.Series): Series of 'volume's
-        length (int): The short period. Default: 13
-        initial (int): The short period. Default: 1000
+        length (int): The short period. Default: 255
+        initial (int): The short period. Default: 100
+        mamode (str): See ``help(ta.ma)``. Default: 'ema'
         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.
+        pd.DataFrame: New DataFrame with ['close', 'volume', 'PVI_1', 'PVIs_<length>']
     """
     # Validate
-    length = v_pos_default(length, 1)
+    mamode = v_mamode(mamode, "ema")
+    length = v_pos_default(length, 255)
     close = v_series(close, length + 1)
     volume = v_series(volume, length + 1)
+    initial = v_pos_default(initial, 100)
+    offset = v_offset(offset)
 
     if close is None or volume is None:
         return
 
-    initial = v_pos_default(initial, 1000)
-    offset = v_offset(offset)
+    # Create a dataframe from the close and volume series
+    # retain index of the close series
+    df = close.to_frame('close')
+    df['volume'] = volume
+
+    df['PVI_1'] = pd.Series(dtype=float)
+    df.iloc[0, df.columns.get_loc('PVI_1')] = initial
+
+    # Get numpy arrays of the data
+    close_prices = df['close'].values
+    volumes = df['volume'].values
+    pvis = np.empty(len(df))
+
+    # Set the first value from from initial
+    pvis[0] = df.iloc[0]['PVI_1']
 
     # Calculate
-    signed_volume = signed_series(volume, 1)
-    _roc = roc(close=close, length=length)
-    pvi = _roc * signed_volume[signed_volume > 0].abs()
-    pvi.fillna(0, inplace=True)
-    pvi.iloc[0] = initial
-    pvi = pvi.cumsum()
-
-    # Offset
+    for i in range(1, len(df)):
+        if volumes[i] > volumes[i-1]:
+            # PVI = Yesterday’s PVI + [[(Close – Yesterday’s Close) / Yesterday’s Close] * Yesterday’s PVI
+            pvis[i] = pvis[i-1] + (((close_prices[i] - close_prices[i-1]) / close_prices[i-1]) * pvis[i-1])
+        else:
+            # PVI = Yesterday’s PVI
+            pvis[i] = pvis[i-1]
+
+    # Update the df
+    df['PVI_1'] = pvis
+    df.name = "PVI_1"
+
     if offset != 0:
-        pvi = pvi.shift(offset)
+        df['PVI_1'] = df['PVI_1'].shift(offset)
+
+    sig_series = ma(mamode, df['PVI_1'], length=length)
+    df[f'PVIs_{length}'] = sig_series
 
     # Fill
     if "fillna" in kwargs:
-        pvi.fillna(kwargs["fillna"], inplace=True)
+        df.fillna(kwargs["fillna"], inplace=True)
     if "fill_method" in kwargs:
-        pvi.fillna(method=kwargs["fill_method"], inplace=True)
-
-    # Name and Category
-    pvi.name = f"PVI_{length}"
-    pvi.category = "volume"
+        df.fillna(method=kwargs["fill_method"], inplace=True)
 
-    return pvi
+    return df

tests/test_indicator_volume.py

@@ -138,7 +138,7 @@ def test_obv(df):
 
 def test_pvi(df):
     result = ta.pvi(df.close, df.volume)
-    assert isinstance(result, Series)
+    assert isinstance(result, DataFrame)
     assert result.name == "PVI_1"
 
 
@@ -253,7 +253,7 @@ def test_ext_nvi(df):
 
 def test_ext_pvi(df):
     df.ta.pvi(append=True)
-    assert df.columns[-1] == "PVI_1"
+    assert list(df.columns) == ["close", "volume", "PVI_1", "PVIs_255"]
 
 
 def test_ext_pvol(df):