Optimize with variable + dataframe column

[jmalizzi]

Hey there friends. I have a fully functioning ML script working with optimization on a variable called "target" but I cannot figure out how to optimize on "target" and a few dataframe columns with ML predictions. I've spent 4 days on this so far. :-) Any help would be appreciated. See screenshots and code below. Columns are RTY_Vol_pred, GC_Close_pred, S_Open_pred.

a single backtest

%%time
from backtesting import Backtest, Strategy
data = df.loc[end_train:].copy()

class ML(Strategy):

target = 1
#column = RTY_Vol_pred

def init(self):        
    0
    
def next(self):       
    #if self.data.column > self.target and not self.position.is_long:
    if self.data.RTY_Vol_pred > self.target and not self.position.is_long:
        
        self.buy()
    
    for trade in self.trades:
        cur_time = self.data.index[-1]
        if cur_time - trade.entry_time >= pd.Timedelta('0 days'):
            trade.close()

bt = Backtest(data, ML, cash=1000000)
bt.run()
#bt.plot()
################################################

optimizer

%%time

from backtesting import Backtest
import numpy as np

backtest = Backtest(data, ML, cash=1000000)

stats, heatmap = backtest.optimize(

target = (.1,.2,.3,.4,.5,.6,.7,.8,.9,1.0,1.1,1.2,1.3,1.4,1.5),
#column = 

#constraint=lambda p: p.n_exit < p.n_enter < p.n1 < p.n2,
#maximize='Equity Final [$]',
maximize='Sharpe Ratio',
max_tries=200,
random_state=0,
return_heatmap=True)

print(heatmap)

[kernc]

What exactly would you like to optimize? We optimize free parameters of the model (logic), not fixed values of target and RTY_Vol_pred which you pass in as data ... 🤔

[jmalizzi]

@kernc i want to optimize through 200+ columns of ML regression predictions i.e. RTY_Vol_pred and a range of target variables which represent the value of the predictions. the 200+ columns are being created before the backtest begins. my example above only shows 3 of the columns as an example.