Fixed bug in df_t

Author: domokane

financepy/market/curves/composite_discount_curve.py

@@ -44,7 +44,7 @@ def __init__(self, child_curves: List[DiscountCurve]):
 
     ###############################################################################
 
-    def _df(self, t: Union[float, np.ndarray]):
+    def df_t(self, t: Union[float, np.ndarray]):
         """
         Return discount factors given a single or vector of dates.
         ParentRate = Sum of children rates => Parent DF = product of children dfs

financepy/market/curves/discount_curve_pwf_onf.py

@@ -22,15 +22,17 @@
 
 
 class DiscountCurvePWFONF(DiscountCurve):
-    """ Curve with piece-wise flat instantaneous (ON) fwd rates. Curve is made up of a series of sections with each having
+    """Curve with piece-wise flat instantaneous (ON) fwd rates. Curve is made up of a series of sections with each having
     a flat instantaneous forward rate. The default compounding assumption is
-    continuous. The class inherits methods from DiscountCurve. """
-
-    def __init__(self,
-                 value_dt: Date,
-                 knot_dates: list,
-                 onfwd_rates: Union[list, np.ndarray]):
-        """ 
+    continuous. The class inherits methods from DiscountCurve."""
+
+    def __init__(
+        self,
+        value_dt: Date,
+        knot_dates: list,
+        onfwd_rates: Union[list, np.ndarray],
+    ):
+        """
         Creates a discount curve using a vector of times and ON fwd rates
         The fwd rate is right-continuous i.e. a given value in the input is applied to the left of that date until the previous date exclusive
         The last fwd rate is extrapolated into the future
@@ -52,25 +54,37 @@ def __init__(self,
         self._freq_type = FrequencyTypes.CONTINUOUS
         self._day_count_type = DayCountTypes.SIMPLE
 
-        dc_times = times_from_dates(self._knot_dates,
-                                    self.value_dt,
-                                    self._day_count_type)
+        dc_times = times_from_dates(
+            self._knot_dates, self.value_dt, self._day_count_type
+        )
 
         self._times = np.atleast_1d(dc_times)
 
         # it is easier to deal in log(dfs), log(df[Ti]) = -\int_0^T_i f(u) du
-        self._logdfs = - \
-            np.cumsum(np.diff(self._times, prepend=0.0) * self._onfwd_rates)
+        self._logdfs = -np.cumsum(
+            np.diff(self._times, prepend=0.0) * self._onfwd_rates
+        )
         self._logdfs_interp = interpolate.interp1d(
-            np.concatenate(([0.0], self._times)), np.concatenate(([0.0], self._logdfs)), kind='linear', bounds_error=False, fill_value='extrapolate')
+            np.concatenate(([0.0], self._times)),
+            np.concatenate(([0.0], self._logdfs)),
+            kind="linear",
+            bounds_error=False,
+            fill_value="extrapolate",
+        )
 
         if test_monotonicity(self._times) is False:
             raise FinError("Times are not sorted in increasing order")
 
     ###############################################################################
 
     @classmethod
-    def brick_wall_curve(cls, valuation_date: Date, start_date: Date, end_date: Date, level: float = 1.0*g_basis_point):
+    def brick_wall_curve(
+        cls,
+        valuation_date: Date,
+        start_date: Date,
+        end_date: Date,
+        level: float = 1.0 * g_basis_point,
+    ):
         """Generate a discount curve of the shape f(t) = level*1_{startdate < t <= enddate} where f(.) is the instantaneous forward rate
             Mostly useful for applying bumps to other discount_curve's, see composite_discount_curve.py
         Args:
@@ -82,23 +96,24 @@ def brick_wall_curve(cls, valuation_date: Date, start_date: Date, end_date: Date
         Returns:
             DiscountCurve: discount curve of the required shape
         """
-        knot_dates = [start_date, end_date, end_date.add_tenor('1D')]
+        knot_dates = [start_date, end_date, end_date.add_tenor("1D")]
         onfwd_rates = [0.0, level, 0.0]
         return cls(valuation_date, knot_dates, onfwd_rates)
 
     ###############################################################################
 
     @classmethod
-    def flat_curve(cls, valuation_date: Date, level: float = 1.0*g_basis_point):
-        knot_dates = [valuation_date.add_tenor('1Y')]
+    def flat_curve(
+        cls, valuation_date: Date, level: float = 1.0 * g_basis_point
+    ):
+        knot_dates = [valuation_date.add_tenor("1Y")]
         onfwd_rates = [level]
         return cls(valuation_date, knot_dates, onfwd_rates)
 
     ###############################################################################
 
-    def _zero_rate(self,
-                   times: Union[float, np.ndarray, list]):
-        """ 
+    def _zero_rate(self, times: Union[float, np.ndarray, list]):
+        """
         Piecewise flat instantaneous (ON) fwd rate is the same as linear logDfs
         """
 
@@ -109,25 +124,23 @@ def _zero_rate(self,
 
         times = np.maximum(times, g_small)
         ldfs = self._logdfs_interp(times)
-        zero_rates = -ldfs/times
+        zero_rates = -ldfs / times
         return zero_rates
 
     ###############################################################################
 
-    def _df(self, t: Union[float, np.ndarray]):
-        """ Return discount factors given a single or vector of times in years. The
+    def df_t(self, t: Union[float, np.ndarray]):
+        """Return discount factors given a single or vector of times in years. The
         discount factor depends on the rate and this in turn depends on its
         compounding frequency and it defaults to continuous compounding. It
         also depends on the day count convention. This was set in the
-        construction of the curve to be ACT_ACT_ISDA. """
+        construction of the curve to be ACT_ACT_ISDA."""
 
         zero_rates = self._zero_rate(t)
 
-        df = self._zero_to_df(self.value_dt,
-                              zero_rates,
-                              t,
-                              self._freq_type,
-                              self._day_count_type)
+        df = self._zero_to_df(
+            self.value_dt, zero_rates, t, self._freq_type, self._day_count_type
+        )
 
         return df
 
@@ -145,7 +158,8 @@ def __repr__(self):
     ###############################################################################
 
     def _print(self):
-        """ Simple print function for backward compatibility. """
+        """Simple print function for backward compatibility."""
         print(self)
 
+
 ###############################################################################