update

README.md

@@ -44,9 +44,10 @@ exits = exits | forced_exits_window
 
 exits.tail(20)
 ```
+## is rising/is falling
+`isrising(series,n)`,`isfalling(series, n)` - returns mask where the condition is satisfied of rising or falling elements including equal values
 
-`isrising(series,n)`,`isfalling(series, n)` - returns mask where the condition is satisfied of consecutive rising or falling elements
+`isrisingc(series,n)`,`isfallingc(series, n)`  - same as above but scritly rising/fallinf (no equal values)
-
 # Indicators
 
 Custom indicators in the `indicators` folder.

setup.py

@@ -2,7 +2,7 @@ from setuptools import setup, find_packages
 
 setup(
     name='ttools',
-    version='0.3.4',
+    version='0.3.5',
     packages=find_packages(),
     install_requires=[
         'vectorbtpro',

ttools/indicators/DIVERGENCE.py

@@ -24,22 +24,28 @@ DIVERGENCE - of two time series, same like in v2realbot
 
 
 @jit(nopython=True)
-def divergence(series1, series2, divtype):
+def divergence(series1, series2, divtype, round):
         #div = a+b   /   a-b  will give value between -1 and 1
         if divtype == "reln":
-            return (series1 - series2) / (series1 + series2)
+            out = (series1 - series2) / (series1 + series2)
         elif divtype == "rel":
-            return series1 - series2
+            out = series1 - series2
         elif divtype == "abs":
-            return np.abs(series1 - series2)
+            out = np.abs(series1 - series2)
         elif divtype == "absn":
-            return np.abs(series1 - series2) / series1
+            out = np.abs(series1 - series2) / series1
         elif divtype == "pctabs":
-            return np.abs(((series1 - series2) / series1) * 100)
+            out = np.abs(((series1 - series2) / series1) * 100)
         elif divtype == "pct":
-            return ((series1 - series2) / series1) * 100
+            out = ((series1 - series2) / series1) * 100
         else:
-            return np.full_like(series1, np.nan)
+            out = np.full_like(series1, np.nan)
+        
+        for i in range(out.shape[0]):
+            if not np.isnan(out[i]):
+                out[i] = np.round(out[i], round)
+
+        return out
 
 """
 Divergence indicator - various divergences between two series
@@ -48,11 +54,12 @@ IND_DIVERGENCE = vbt.IF(
     class_name='DIVERGENCE',
     module_name='ttools',
     input_names=['series1', 'series2'],
-    param_names=["divtype"],
+    param_names=["divtype", "round"],
     output_names=['div']
 ).with_apply_func(divergence,
                 takes_1d=True,
                 param_settings=dict(
                 ),
-                divtype="reln"
+                divtype="reln",
+                round=4
                 )

ttools/vbtutils.py

@@ -4,6 +4,7 @@ import pandas_market_calendars as mcal
 from typing import Any
 import datetime
 
+#TBD create NUMBA alternatives
 def isrising(series: pd.Series, n: int) -> pd.Series:
     """
     Checks if a series is rising over a given window size.
@@ -23,6 +24,32 @@ def isrising(series: pd.Series, n: int) -> pd.Series:
     """
     return series.rolling(n).apply(lambda x: (x == sorted(x, reverse=False)).all(), raw=False).fillna(False).astype(bool)
 
+def isrisingc(series: pd.Series, n: int) -> pd.Series:
+    """
+    Checks if a series is strictly rising over a given window size.
+    Returns True for windows where values are strictly increasing.
+    
+    Parameters
+    ----------
+    series : pd.Series
+        Input series
+    n : int
+        Window size
+        
+    Returns
+    -------
+    pd.Series
+        Boolean mask indicating when the series is strictly rising
+    """
+    # Calculate the difference between consecutive values
+    diffs = series.diff()
+
+    # We check if all values in the window are negative (falling)
+    result = diffs.rolling(n-1).apply(lambda x: (x > 0).all(), raw=True)
+    
+    # Fill the first n-1 values with False and return the boolean mask
+    return result.fillna(False).astype(bool)
+
 def isfalling(series: pd.Series, n: int) -> pd.Series:
     """
     Checks if a series is falling over a given window size.
@@ -41,6 +68,33 @@ def isfalling(series: pd.Series, n: int) -> pd.Series:
     """
     return series.rolling(n).apply(lambda x: (x == sorted(x, reverse=True)).all(), raw=False).fillna(False).astype(bool)
 
+def isfallingc(series: pd.Series, n: int) -> pd.Series:
+    """
+    Checks if a series is strictly falling over a given window size.
+    Returns True for windows where values are strictly decreasing.
+    
+    Parameters
+    ----------
+    series : pd.Series
+        Input series
+    n : int
+        Window size
+        
+    Returns
+    -------
+    pd.Series
+        Boolean mask indicating when the series is strictly falling
+    """
+    # Calculate the difference between consecutive values
+    diffs = series.diff()
+
+    # We check if all values in the window are negative (falling)
+    result = diffs.rolling(n-1).apply(lambda x: (x < 0).all(), raw=True)
+    
+    # Fill the first n-1 values with False and return the boolean mask
+    return result.fillna(False).astype(bool)
+
+
 def create_mask_from_window(series: Any, entry_window_opens:int, entry_window_closes:int, use_cal: bool = True):
     """
     Accepts series and window range (number of minutes from market start) and returns boolean mask denoting