Fixed various deprecation messages related Pandas

Author: saeedamen

README.md

@@ -72,7 +72,7 @@ https://github.com/cuemacro/finmarketpy/blob/master/INSTALL.md (which includes d
 
 Also take a look at https://github.com/cuemacro/teaching/blob/master/pythoncourse/installation/installing_anaconda_and_pycharm.ipynb
 from my Python for finance workshop course, where I keep notes specifically about setting up your Anaconda environment
-for data science (including for findatapy/chartpy/finmarketpy), including YAML files etc.
+for data science (including for findatapy/chartpy/finmarketpy) etc.
 
 You can install the library using the below (better to get the newest version from repo, as opposed to releases).
 
@@ -102,12 +102,15 @@ pip install chartpy
 pip install findatapy
 ```
 
-Note that if you use the option pricing/total returns you might need to get the latest FinancePy version from GitHub
-https://github.com/domokane/FinancePy/ as opposed to PyPI
+FinancePy is an optional dependency for finmarketpy for option pricing. It is recommended
+to install it separately from PyPI after installing finmarketpy, and without dependencies
+otherwise it can cause clashes with other libraries (because of its strict version
+dependencies on libraries like llvmlite, which in practice can be relaxed). 
+The API changes a lot so it recommended to install the specific version listed below.
 
 ```
-pip install git+https://github.com/domokane/FinancePy/FinancePy.git
-
+pip install numba numpy scipy llvmlite ipython pandas prettytable
+pip install financepy==0.370 --no-deps
 ```
 
 # Binder and Jupyter - Run finmarketpy in your browser
@@ -158,17 +161,19 @@ You may sometimes experience Numba errors like such as `Failed in nopython mode
 
 One possible way to fix this is to delete the `__pycache__` folders underneath wherever financepy is installed:
 
-Eg. if you are using the `py38class` environment, if you've installed Anaconda in `C:\Anaconda3`, you might find the financepy
+Eg. if you are using the `py310class` environment, if you've installed Anaconda in `C:\Anaconda3`, you might find the financepy
 folder at the below location
 
-`C:\Anaconda3\envs\py38class\Lib\site-packages\financepy`
+`C:\Anaconda3\envs\py310class\Lib\site-packages\financepy`
 
 # finmarketpy examples
 
 In finmarketpy/examples you will find several examples, including some simple trading models
 
 # Release Notes
 
+* 0.11.14 - finmarketpy (08 Mar 2025)
+* 0.11.13 - finmarketpy (01 Jan 2024)
 * 0.11.12 - finmarketpy (26 Apr 2023)
 * 0.11.11 - finmarketpy (07 Oct 2021)
 * 0.11.10 - finmarketpy (06 Oct 2021)
@@ -187,6 +192,10 @@ In finmarketpy/examples you will find several examples, including some simple tr
 
 # finmarketpy log
 
+* 08 Mar 2025
+  * Make FinancePy an optional dependency
+  * Fixed various deprecation messages related Pandas
+  * Made Plotly charts look nicer (autoscale with latest ChartPy)
 * 07 Mar 2025
   * Merge changes for pyproject.toml etc.
   * Formatting towards PEP8

finmarketpy/backtest/backtestengine.py

@@ -187,7 +187,7 @@ def calculate_trading_PnL(
         # Only allow signals to change on the days when we can trade assets
         signal_df = signal_df.mask(
             non_trading_days)  # fill asset holidays with NaN signals
-        signal_df = signal_df.fillna(method='ffill')  # fill these down
+        signal_df = signal_df.ffill()  # fill these down
 
         # Transaction costs and roll costs
         tc = br.spot_tc_bp
@@ -202,8 +202,7 @@ def calculate_trading_PnL(
             pnl_cols.append(asset_df_cols[i] + " / " + signal_cols[i])
 
         # Fill down asset holidays (we won't trade on these days)
-        asset_df = asset_df.fillna(
-            method='ffill')
+        asset_df = asset_df.ffill()
         returns_df = calculations.calculate_returns(asset_df)
 
         # Apply a stop loss/take profit to every trade if this has been specified
@@ -969,6 +968,7 @@ class TradingModel(object):
     HEIGHT = ChartConstants().chartfactory_height
     CHART_SOURCE = ChartConstants().chartfactory_source
     CHART_STYLE = Style()
+    AUTO_SCALE = True
     PLOTLY_PLOT_MODE = "offline_html_exc_embed_js"
 
     DUMP_CSV = ''
@@ -1428,7 +1428,7 @@ def compare_strategy_vs_benchmark(
             # Only calculate return statistics if this has been specified (note when different frequencies of data
             # might underrepresent vol
             # if calc_stats:
-            benchmark_df = benchmark_df.fillna(method='ffill')
+            benchmark_df = benchmark_df.ffill()
             benchmark_df = self._filter_by_plot_start_finish_date(benchmark_df,
                                                                   br)
 
@@ -1440,8 +1440,7 @@ def compare_strategy_vs_benchmark(
 
             # Realign strategy & benchmark
             strategy_benchmark_df = strategy_df.join(benchmark_df, how='inner')
-            strategy_benchmark_df = strategy_benchmark_df.fillna(
-                method='ffill')
+            strategy_benchmark_df = strategy_benchmark_df.ffill()
 
             strategy_benchmark_df = self._filter_by_plot_start_finish_date(
                 strategy_benchmark_df, br)
@@ -1623,7 +1622,7 @@ def _reduce_plot(self,
             if reduce_plot and resample is not None:
                 # make plots on every business day (will downsample intraday data)
                 data_frame = data_frame.resample(resample).last()
-                data_frame = data_frame.fillna(method='pad')
+                data_frame = data_frame.ffill(method='pad')
 
             return data_frame
         except:
@@ -2017,13 +2016,13 @@ def _plot_ret_stats_helper(self, ret_stats, metric, title, file_tag,
 
         for key in keys:
             if metric == "IR":
-                ret_metric.append(ret_stats[key].inforatio()[0])
+                ret_metric.append(ret_stats[key].inforatio().iloc[0])
             elif metric == "Returns":
-                ret_metric.append(ret_stats[key].ann_returns()[0] * 100)
+                ret_metric.append(ret_stats[key].ann_returns().iloc[0] * 100)
             elif metric == "Vol":
-                ret_metric.append(ret_stats[key].ann_vol()[0] * 100)
+                ret_metric.append(ret_stats[key].ann_vol().iloc[0] * 100)
             elif metric == "Drawdowns":
-                ret_metric.append(ret_stats[key].drawdowns()[0] * 100)
+                ret_metric.append(ret_stats[key].drawdowns().iloc[0] * 100)
 
         if strip is not None: keys = [k.replace(strip, '') for k in keys]
 
@@ -2113,7 +2112,8 @@ def plot_strategy_group_leverage(
                                           ret_with_df=ret_with_df,
                                           split_on_char=split_on_char)
 
-    ###### Plot signals and trades, in terms of units, notionals and contract sizes (eg. for futures)
+    ###### Plot signals and trades, in terms of units, notionals and
+    # contract sizes (eg. for futures)
 
     def plot_strategy_all_signals(
             self,
@@ -2483,6 +2483,9 @@ def _create_style(self, title, file_add, reduce_plot=True):
         style.source = self.CHART_SOURCE
         style.silent_display = not (self.SHOW_CHARTS)
         style.plotly_plot_mode = self.PLOTLY_PLOT_MODE  # Smaller file sizes
+        style.auto_scale = self.AUTO_SCALE
+
+        print(style.plotly_plot_mode)
 
         style.legend_bgcolor = 'rgba(0,0,0,0)'
 

finmarketpy/backtest/tradeanalysis.py

@@ -52,7 +52,7 @@ def __init__(self, engine=ChartConstants().chartfactory_default_engine):
 
         return
 
-    def run_strategy_returns_stats(self, trading_model):
+    def run_strategy_returns_stats(self, trading_model, engine="finmarketpy"):
         """Plots useful statistics for the trading strategy using various backends
 
         Parameters
@@ -335,8 +335,8 @@ def run_arbitrary_sensitivity(self, trading_model, parameter_list=None,
 
         style = Style()
 
-        ir = [t.inforatio()[0] for t in ret_stats_list]
-        rets = [t.ann_returns()[0] for t in ret_stats_list]
+        ir = [t.inforatio().iloc[0] for t in ret_stats_list]
+        rets = [t.ann_returns().iloc[0] for t in ret_stats_list]
 
         # if we have too many combinations remove legend and use scaled shaded colour
         # if len(port_list) > 10:

finmarketpy/curve/fxoptionscurve.py

@@ -792,10 +792,10 @@ def apply_tc_signals_to_total_return_index(self, cross_fx,
                       total_return_index_df[cross + '-delta.close'].shift(
                           1)) * spot_tc
 
-            total_return_index_df[cross + '-option-return-with-tc.close'][
-                0] = 0
-            total_return_index_df[cross + '-delta-pnl-return-with-tc.close'][
-                0] = 0
+            total_return_index_df.iloc[0,
+                total_return_index_df.columns.get_loc(cross + '-option-return-with-tc.close')] = 0
+            total_return_index_df.iloc[0,
+                total_return_index_df.columns.get_loc(cross + '-delta-pnl-return-with-tc.close')] = 0
             total_return_index_df[
                 cross + '-option-delta-return-with-tc.close'] = \
                 total_return_index_df[cross + '-option-return-with-tc.close'] + \

finmarketpy/curve/fxspotcurve.py

@@ -260,11 +260,11 @@ def construct_total_return_index(self, cross_fx, market_df,
 
                 # Sometimes depo data can be patchy, ok to fill down, given not
                 # very volatile (don't do this with spot!)
-                carry = carry.fillna(method='ffill') / 100.0
+                carry = carry.ffill() / 100.0
 
                 # In case there are values missing at start of list (fudge for
                 # old data!)
-                carry = carry.fillna(method='bfill')
+                carry = carry.bfill()
 
                 spot = spot[cross + "." + field].to_frame()
 
@@ -274,11 +274,12 @@ def construct_total_return_index(self, cross_fx, market_df,
                 terms_deposit_vals = carry[
                     cross[3:6] + depo_tenor + "." + field].values
 
-                # Calculate the time difference between each data point (flooring it to whole days, because carry
+                # Calculate the time difference between each data point (
+                # flooring it to whole days, because carry
                 # is accured when there's a new day)
                 spot['index_col'] = spot.index.floor('D')
                 time = spot['index_col'].diff()
-                spot = spot.drop('index_col', 1)
+                spot = spot.drop('index_col', axis=1)
 
                 time_diff = time.values.astype(
                     float) / 86400000000000.0  # get time difference in days

finmarketpy/curve/volatility/fxvolsurface.py

@@ -32,7 +32,7 @@
         FinFXDeltaMethod
     from financepy.market.volatility.fx_vol_surface_plus import vol_function
     from financepy.market.volatility.fx_vol_surface_plus import \
-        VolFunctionTypes
+        VolFuncTypes
 
     from financepy.utils.global_types import FinSolverTypes
 except:
@@ -142,17 +142,17 @@ def __init__(self, market_df=None, asset=None, field="close",
             [asset + "10R" + t + field for t in tenors]].values
 
         if vol_function_type == "CLARK":
-            self._vol_function_type = VolFunctionTypes.CLARK
+            self._vol_function_type = VolFuncTypes.CLARK
         elif vol_function_type == "CLARK5":
-            self._vol_function_type = VolFunctionTypes.CLARK5
+            self._vol_function_type = VolFuncTypes.CLARK5
         elif vol_function_type == "BBG":
-            self._vol_function_type = VolFunctionTypes.BBG
+            self._vol_function_type = VolFuncTypes.BBG
 
         # Note: currently SABR isn"t fully implemented in FinancePy
         elif vol_function_type == "SABR":
-            self._vol_function_type = VolFunctionTypes.SABR
+            self._vol_function_type = VolFuncTypes.SABR
         elif vol_function_type == "SABR3":
-            self._vol_function_type = VolFunctionTypes.SABR3
+            self._vol_function_type = VolFuncTypes.SABR3
 
         # What does ATM mean? (for most
         if atm_method == "fwd-delta-neutral":  # ie. strike such that a straddle would be delta neutral
@@ -246,10 +246,10 @@ def build_vol_surface(self, value_date):
             self._risk_reversal10DeltaVols[
                 date_index][0],
             self._alpha,
-            atmMethod=self._atm_method,
-            deltaMethod=self._delta_method,
-            volatility_function_type=self._vol_function_type,
-            finSolverType=self._solver,
+            atm_method=self._atm_method,
+            delta_method=self._delta_method,
+            vol_func_type=self._vol_function_type,
+            fin_solver_type=self._solver,
             tol=self._tol)  # TODO add tol
 
     def calculate_vol_for_strike_expiry(self, K, expiry_date=None, tenor="1M"):
@@ -335,21 +335,21 @@ def extract_vol_surface(self, num_strike_intervals=60, low_K_pc=0.95,
 
         # columns = tenors
         df_vol_surface_strike_space = pd.DataFrame(
-            columns=self._fin_fx_vol_surface._tenors)
+            columns=self._fin_fx_vol_surface.tenors)
         df_vol_surface_delta_space = pd.DataFrame(
-            columns=self._fin_fx_vol_surface._tenors)
+            columns=self._fin_fx_vol_surface.tenors)
 
         # columns = tenors
         df_vol_surface_implied_pdf = pd.DataFrame(
-            columns=self._fin_fx_vol_surface._tenors)
+            columns=self._fin_fx_vol_surface.tenors)
 
         # Conversion between main deltas and strikes
         df_deltas_vs_strikes = pd.DataFrame(
-            columns=self._fin_fx_vol_surface._tenors)
+            columns=self._fin_fx_vol_surface.tenors)
 
         # ATM, 10d + 25d market strangle and 25d risk reversals
         df_vol_surface_quoted_points = pd.DataFrame(
-            columns=self._fin_fx_vol_surface._tenors)
+            columns=self._fin_fx_vol_surface.tenors)
 
         # Note, at present we"re not using 10d strikes
         quoted_strikes_names = ["ATM", "STR_25D_MS", "RR_25D_P", "STR_10D_MS",
@@ -359,8 +359,8 @@ def extract_vol_surface(self, num_strike_intervals=60, low_K_pc=0.95,
                              "K_10D_C_MS"]
 
         # Get max/min strikes to interpolate (from the longest dated tenor)
-        low_K = self._fin_fx_vol_surface._K_25D_P[-1] * low_K_pc
-        high_K = self._fin_fx_vol_surface._K_25D_C[-1] * high_K_pc
+        low_K = self._fin_fx_vol_surface.k_25d_p[-1] * low_K_pc
+        high_K = self._fin_fx_vol_surface.k_25d_c[-1] * high_K_pc
 
         if num_strike_intervals is not None:
             # In case using old version of FinancePy
@@ -370,19 +370,19 @@ def extract_vol_surface(self, num_strike_intervals=60, low_K_pc=0.95,
             except:
                 pass
 
-        for tenor_index in range(0, self._fin_fx_vol_surface._num_vol_curves):
+        for tenor_index in range(0, self._fin_fx_vol_surface.num_vol_curves):
 
             # Get the quoted vol points
-            tenor_label = self._fin_fx_vol_surface._tenors[tenor_index]
+            tenor_label = self._fin_fx_vol_surface.tenors[tenor_index]
 
-            atm_vol = self._fin_fx_vol_surface._atm_vols[tenor_index] * 100
-            ms_25d_vol = self._fin_fx_vol_surface._mktStrangle25DeltaVols[
+            atm_vol = self._fin_fx_vol_surface.atm_vols[tenor_index] * 100
+            ms_25d_vol = self._fin_fx_vol_surface.ms_25_delta_vols[
                              tenor_index] * 100
-            rr_25d_vol = self._fin_fx_vol_surface._riskReversal25DeltaVols[
+            rr_25d_vol = self._fin_fx_vol_surface.rr_25_delta_vols[
                              tenor_index] * 100
-            ms_10d_vol = self._fin_fx_vol_surface._mktStrangle10DeltaVols[
+            ms_10d_vol = self._fin_fx_vol_surface.ms_10_delta_vols[
                              tenor_index] * 100
-            rr_10d_vol = self._fin_fx_vol_surface._riskReversal10DeltaVols[
+            rr_10d_vol = self._fin_fx_vol_surface.rr_10_delta_vols[
                              tenor_index] * 100
 
             df_vol_surface_quoted_points[tenor_label] = pd.Series(
@@ -416,19 +416,19 @@ def extract_vol_surface(self, num_strike_intervals=60, low_K_pc=0.95,
 
             # Extract strikes for the quoted points (ie. 10d, 25d and ATM)
             key_strikes = []
-            key_strikes.append(self._fin_fx_vol_surface._K_10D_P[tenor_index])
+            key_strikes.append(self._fin_fx_vol_surface.k_10d_p[tenor_index])
             key_strikes.append(
-                self._fin_fx_vol_surface._K_10D_P_MS[tenor_index])
-            key_strikes.append(self._fin_fx_vol_surface._K_25D_P[tenor_index])
+                self._fin_fx_vol_surface.k_10d_p_ms[tenor_index])
+            key_strikes.append(self._fin_fx_vol_surface.k_25d_p[tenor_index])
             key_strikes.append(
-                self._fin_fx_vol_surface._K_25D_P_MS[tenor_index])
-            key_strikes.append(self._fin_fx_vol_surface._K_ATM[tenor_index])
-            key_strikes.append(self._fin_fx_vol_surface._K_25D_C[tenor_index])
+                self._fin_fx_vol_surface.k_25d_p_ms[tenor_index])
+            key_strikes.append(self._fin_fx_vol_surface.k_atm[tenor_index])
+            key_strikes.append(self._fin_fx_vol_surface.k_25d_c[tenor_index])
             key_strikes.append(
-                self._fin_fx_vol_surface._K_25D_C_MS[tenor_index])
-            key_strikes.append(self._fin_fx_vol_surface._K_10D_C[tenor_index])
+                self._fin_fx_vol_surface.k_25d_c_ms[tenor_index])
+            key_strikes.append(self._fin_fx_vol_surface.k_10d_c[tenor_index])
             key_strikes.append(
-                self._fin_fx_vol_surface._K_10D_C_MS[tenor_index])
+                self._fin_fx_vol_surface.k_10d_c_ms[tenor_index])
 
             df_deltas_vs_strikes[tenor_label] = pd.Series(
                 index=key_strikes_names, data=key_strikes)
@@ -467,9 +467,9 @@ def get_vol_from_quoted_tenor(self, K, tenor, gaps=None):
         if gaps is None:
             gaps = np.array([0.1])
 
-        params = self._fin_fx_vol_surface._parameters[tenor_index]
-        t = self._fin_fx_vol_surface._texp[tenor_index]
-        f = self._fin_fx_vol_surface._F0T[tenor_index]
+        params = self._fin_fx_vol_surface.parameters[tenor_index]
+        t = self._fin_fx_vol_surface.t_exp[tenor_index]
+        f = self._fin_fx_vol_surface.fwd[tenor_index]
 
         return vol_function(self._vol_function_type.value, params,
                             np.array([K]), gaps, f, K, t)

finmarketpy/economics/seasonality.py

@@ -149,7 +149,7 @@ def monthly_seasonality(self, data_frame,
 
         if price_index:
             data_frame = data_frame.resample(
-                'BM').mean()  # resample into month end
+                'BME').mean()  # resample into month end
             data_frame = calculations.calculate_returns(data_frame)
 
         data_frame.index = pandas.to_datetime(data_frame.index)
@@ -209,7 +209,7 @@ def _remove_seasonality(self, series, likely_period=None):
         if adjusted is None:
             return numpy.nan
 
-        return adjusted[-1]
+        return adjusted.iloc[-1]
 
 
 if __name__ == '__main__':