Optimize with blackbox¶

Home

Install latest blackbox optimize from PyPi

In [ ]:

!pip install git+https://github.com/paulknysh/blackbox.git@master

!pip install git+https://github.com/paulknysh/blackbox.git@master

Sample strategy¶

In [1]:

import talib.abstract as ta

from lettrade import DataFeed, Strategy, indicator as i
from lettrade.exchange.backtest import ForexBackTestAccount, let_backtest


class SmaCross(Strategy):
    ema1_window = 9
    ema2_window = 21

    def indicators(self, df: DataFeed):
        df["ema1"] = ta.EMA(df, timeperiod=self.ema1_window)
        df["ema2"] = ta.EMA(df, timeperiod=self.ema2_window)

        df["signal_ema_crossover"] = i.crossover(df.ema1, df.ema2)
        df["signal_ema_crossunder"] = i.crossunder(df.ema1, df.ema2)

    def next(self, df: DataFeed):
        if len(self.orders) > 0 or len(self.positions) > 0:
            return

        if df.l.signal_ema_crossover[-1]:
            price = self.data.l.close[-1]
            self.buy(size=0.1, sl=price - 0.001, tp=price + 0.001)
        elif df.l.signal_ema_crossunder[-1]:
            price = self.data.l.close[-1]
            self.sell(size=0.1, sl=price + 0.001, tp=price - 0.001)


lt = let_backtest(
    strategy=SmaCross,
    datas="example/data/data/EURUSD_5m-0_10000.csv",
    account=ForexBackTestAccount,
)

import talib.abstract as ta from lettrade import DataFeed, Strategy, indicator as i from lettrade.exchange.backtest import ForexBackTestAccount, let_backtest class SmaCross(Strategy): ema1_window = 9 ema2_window = 21 def indicators(self, df: DataFeed): df["ema1"] = ta.EMA(df, timeperiod=self.ema1_window) df["ema2"] = ta.EMA(df, timeperiod=self.ema2_window) df["signal_ema_crossover"] = i.crossover(df.ema1, df.ema2) df["signal_ema_crossunder"] = i.crossunder(df.ema1, df.ema2) def next(self, df: DataFeed): if len(self.orders) > 0 or len(self.positions) > 0: return if df.l.signal_ema_crossover[-1]: price = self.data.l.close[-1] self.buy(size=0.1, sl=price - 0.001, tp=price + 0.001) elif df.l.signal_ema_crossunder[-1]: price = self.data.l.close[-1] self.sell(size=0.1, sl=price + 0.001, tp=price - 0.001) lt = let_backtest( strategy=SmaCross, datas="example/data/data/EURUSD_5m-0_10000.csv", account=ForexBackTestAccount, )

Run optimize¶

In [2]:

import blackbox as bb


def params_parser(args):
    return {"ema1_window": int(args[0]), "ema2_window": int(args[1])}


def result_parser(result):
    return result["equity"]


result = bb.minimize(
    f=lt.optimize_model(
        params_parser=params_parser,
        result_parser=result_parser,
    ),
    domain=[[5, 25, 1], [10, 50, 1]],  # ranges of each parameter
    budget=300,  # total number of function calls available
    batch=12,  # number of calls that will be evaluated in parallel
)

lt.optimize_done()

import blackbox as bb def params_parser(args): return {"ema1_window": int(args[0]), "ema2_window": int(args[1])} def result_parser(result): return result["equity"] result = bb.minimize( f=lt.optimize_model( params_parser=params_parser, result_parser=result_parser, ), domain=[[5, 25, 1], [10, 50, 1]], # ranges of each parameter budget=300, # total number of function calls available batch=12, # number of calls that will be evaluated in parallel ) lt.optimize_done()

INFO     evaluating batch 1/25 (samples 1..12/300) 07-02 16:52:35
INFO     evaluating batch 2/25 (samples 13..24/300) 07-02 16:52:36
INFO     evaluating batch 3/25 (samples 25..36/300) 07-02 16:52:36
INFO     evaluating batch 4/25 (samples 37..48/300) 07-02 16:52:37
INFO     evaluating batch 5/25 (samples 49..60/300) 07-02 16:52:37
INFO     evaluating batch 6/25 (samples 61..72/300) 07-02 16:52:37
INFO     evaluating batch 7/25 (samples 73..84/300) 07-02 16:52:38
INFO     evaluating batch 8/25 (samples 85..96/300) 07-02 16:52:38
INFO     evaluating batch 9/25 (samples 97..108/300) 07-02 16:52:39
INFO     evaluating batch 10/25 (samples 109..120/300) 07-02 16:52:39
INFO     evaluating batch 11/25 (samples 121..132/300) 07-02 16:52:40
INFO     evaluating batch 12/25 (samples 133..144/300) 07-02 16:52:40
INFO     evaluating batch 13/25 (samples 145..156/300) 07-02 16:52:41
INFO     evaluating batch 14/25 (samples 157..168/300) 07-02 16:52:41
INFO     evaluating batch 15/25 (samples 169..180/300) 07-02 16:52:44
INFO     evaluating batch 16/25 (samples 181..192/300) 07-02 16:52:46
INFO     evaluating batch 17/25 (samples 193..204/300) 07-02 16:52:48
INFO     evaluating batch 18/25 (samples 205..216/300) 07-02 16:52:50
INFO     evaluating batch 19/25 (samples 217..228/300) 07-02 16:52:52
INFO     evaluating batch 20/25 (samples 229..240/300) 07-02 16:52:54
INFO     evaluating batch 21/25 (samples 241..252/300) 07-02 16:52:57
INFO     evaluating batch 22/25 (samples 253..264/300) 07-02 16:53:00
INFO     evaluating batch 23/25 (samples 265..276/300) 07-02 16:53:03
INFO     evaluating batch 24/25 (samples 277..288/300) 07-02 16:53:06
INFO     evaluating batch 25/25 (samples 289..300/300) 07-02 16:53:09
INFO     DONE 07-02 16:53:13

Plot¶

In [3]:

lt.plotter.heatmap(x="ema1_window", y="ema2_window")

lt.plotter.heatmap(x="ema1_window", y="ema2_window")

In [4]:

lt.plotter.contour(x="ema1_window", y="ema2_window")

lt.plotter.contour(x="ema1_window", y="ema2_window")

Plot Plotly¶

Init Plotly environment¶

In [5]:

import plotly.io as pio

pio.renderers.default = "notebook"
pio.templates.default = "plotly_dark"

import plotly.io as pio pio.renderers.default = "notebook" pio.templates.default = "plotly_dark"

Clean data¶

In [6]:

import pandas as pd

df = pd.DataFrame(
    {
        "ema1_window": [r[0] for r in result["all_xs"]],
        "ema2_window": [r[1] for r in result["all_xs"]],
        "equity": result["all_fs"],
    }
)

df.columns = ["x", "y", "z"]
df.x = df.x.astype(int)
df.y = df.y.astype(int)

df

import pandas as pd df = pd.DataFrame( { "ema1_window": [r[0] for r in result["all_xs"]], "ema2_window": [r[1] for r in result["all_xs"]], "equity": result["all_fs"], } ) df.columns = ["x", "y", "z"] df.x = df.x.astype(int) df.y = df.y.astype(int) df

Out[6]:

	x	y	z
0	10	12	9833.68
1	5	35	9874.68
2	20	18	10074.90
3	15	41	9981.80
4	10	23	9857.78
...	...	...	...
295	10	28	9845.10
296	21	34	9890.22
297	6	39	9905.94
298	10	14	9809.16
299	16	32	9927.24

300 rows × 3 columns

Type 1¶

In [7]:

import plotly.express as px

fig = px.density_contour(
    df,
    x="x",
    y="y",
    z="z",
    histfunc="max",
)
fig.update_traces(contours_coloring="fill", contours_showlabels=True)
fig.show()

import plotly.express as px fig = px.density_contour( df, x="x", y="y", z="z", histfunc="max", ) fig.update_traces(contours_coloring="fill", contours_showlabels=True) fig.show()

Type 2¶

In [8]:

import plotly.express as px

fig = px.density_heatmap(
    df,
    x="x",
    y="y",
    z="z",
    nbinsx=20,
    nbinsy=40,
    histfunc="max",
    color_continuous_scale="Viridis",
)
fig.show()

import plotly.express as px fig = px.density_heatmap( df, x="x", y="y", z="z", nbinsx=20, nbinsy=40, histfunc="max", color_continuous_scale="Viridis", ) fig.show()

Type 3¶

In [9]:

import plotly.graph_objects as go

fig = go.Figure(
    go.Histogram2d(
        x=df.x,
        y=df.y,
        z=df.z,
        nbinsx=20,
        nbinsy=40,
        histfunc="max",
    )
)

fig.show()

import plotly.graph_objects as go fig = go.Figure( go.Histogram2d( x=df.x, y=df.y, z=df.z, nbinsx=20, nbinsy=40, histfunc="max", ) ) fig.show()

Source¶

--8<-- "example/optimize/blackbox.py"