stock_strategy.py

To sell only the stocks that are not there in this week's new strategy and keep the stocks that still remain, you can modify the code as follows:

```python
import backtrader as bt
import pandas as pd

# Define the strategy
class MyStrategy(bt.Strategy):
    
    def __init__(self):
        self.market_cap = self.datas[0].market_cap
        self.pb_ratio = self.datas[0].pb_ratio
        self.debt_to_equity = self.datas[0].debt_to_equity
        self.eps_latest_quarter = self.datas[0].eps_latest_quarter
        self.eps_last_year = self.datas[0].eps_last_year
        self.pat_last_year = self.datas[0].pat_last_year
        self.volume_1week_avg = self.datas[0].volume_1week_avg
        
        self.weekly_sma = bt.indicators.SimpleMovingAverage(self.data.close, period=5)
        self.weekly_rsi = bt.indicators.RelativeStrengthIndex(self.data.close, period=14)
        
        self.weekly_buy_signal = False
        self.weekly_stocks = []
        
    def next(self):
        if self.datas[0].datetime.date(0).weekday() == 0:
            # Every Monday
            self.weekly_buy_signal = False
            self.weekly_stocks = []
            
            if self.market_cap[0] > 300 and self.pb_ratio[0] < self.data.close[0] and \
               self.debt_to_equity[0] < 0.25 and self.eps_latest_quarter[0] > self.eps_latest_quarter[-1] and \
               self.eps_last_year[0] > self.eps_last_year[-1] and self.pat_last_year[0] > 0 and \
               self.volume_1week_avg[0] > 700000 and self.weekly_rsi[0] > 50:
                self.weekly_buy_signal = True
                self.weekly_stocks = [data._name for data in self.datas if self.market_cap[data]._value[0] > 300 and 
                                      self.pb_ratio[data]._value[0] < self.data.close[data]._value[0] and 
                                      self.debt_to_equity[data]._value[0] < 0.25 and 
                                      self.eps_latest_quarter[data]._value[0] > self.eps_latest_quarter[data]._value[-1] and 
                                      self.eps_last_year[data]._value[0] > self.eps_last_year[data]._value[-1] and 
                                      self.pat_last_year[data]._value[0] > 0 and 
                                      self.volume_1week_avg[data]._value[0] > 700000]
        
        for data in self.datas:
            if data._name not in self.weekly_stocks and self.getposition(data).size != 0:
                self.sell(data=data, size=self.getposition(data).size)
                
        if self.weekly_buy_signal:
            cash = self.broker.getcash()
            if cash > 0:
                target_value = cash / 15
                size = int(target_value / self.data.close[0])
                for data in self.datas:
                    if data._name in self.weekly_stocks and self.getposition(data).size == 0:
                        self.buy(data=data, size=size)

# Load the data
data = bt.feeds.GenericCSVData(
    dataname='data.csv',
    fromdate=pd.Timestamp('2010-01-01'),
    todate=pd.Timestamp('2021-12-31'),
    nullvalue=0.0,
    dtformat='%Y-%m-%d',
    datetime=0,
    open=1,
    high=2,
    low=3,
    close=4,
    volume=5,
    openinterest=-1,
    market_cap=6,
    pb_ratio=7,
    debt_to_equity=8,
    eps_latest_quarter=9,
    eps_last_year=10,
    pat_last_year=11,
    volume_1week_avg=12
)

# Create the cerebro instance
cerebro = bt.Cerebro()

# Add the data feed
cerebro.adddata(data)

# Add the strategy
cerebro.addstrategy(MyStrategy)

# Set the initial capital
cerebro.broker.setcash
# Set the commission
cerebro.broker.setcommission(commission=0.001)

# Run the backtest
cerebro.run()

# Print the final portfolio value
print('Final Portfolio Value: %.2f' % cerebro.broker.getvalue())