如何将在函数外部、循环中定义的变量输入到函数中？

Question

I'm backtesting some trading strategies in pandas and I defined the function:

def TakeOutDuplicates(DataFrame):
    Buy = []
    Sell = []
    flag = -1
    for i in range(len(DataFrame)-1):
        if DataFrame[ (takeOutThisStrategys[x][Time][j+1]) ][i] == "Buy":
            Sell.append(np.nan)
            if flag != 1:
                Buy.append(DataFrame["Open"][i+1])
                flag=1
            else: 
                Buy.append(np.nan)
        elif DataFrame[ (takeOutThisStrategys[x][Time][j+1]) ][i] == "Sell":
            Buy.append(np.nan)
            if flag != 0:
                Sell.append(DataFrame["Open"][i+1])
                flag=0
            else: 
                Sell.append(np.nan)
        else:
            Buy.append(np.nan)
            Sell.append(np.nan)
    return(Buy, Sell)

Which searches through specific columns in my dataframe and reduces the buy and sell signals into a series so I can analyse what prices it bought at and what prices it sold at, rather than seeing it told me to buy 30 times in a row, it just records the buy price at the first time it said buy, and then says the sell price at the time it said to sell.

The loop I'm running to call this function is:

for Time in TimePeriods:  # TimePeriods = [5, 15, 60, 240, 1440] (in minutes)
    for j in range(savedi-1):      # savedi = 15
        takeOutThisStrategys[x][Time][j+1] = f"Strategy {j+1} on {Time} Mins"
        replaces[x][Time][j+1] = TakeOutDuplicates(df)

As you can see I want to populate my dictionary replaces with the buy and sell prices as returned from the function TakeOutDuplicates. Except when I run this and check my dictionary like the following:

Checking the 5 minute timeframe:

replaces[x][5]

{1: ([14185.11,
   nan,
   nan,
   14185.03,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   14121.72,...])}

Now checking the 240 min (4 hour) time frame.

replaces[x][240]

{1: ([14185.11,
   nan,
   nan,
   14185.03,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   nan,
   14121.72,...])}

As you can see all the buy and sell signals are the same, which is not how it should be, implying that when i run my loop, it's not taking the variables for Time and j and then inputting them into my function, as I would have expected. How can I solve this issue?

Answer 1

I managed to fix this issue, but the problem was in the fact that when I defined my strategies I forgot to multiply by my time periods:

So I had this:

df.loc[( (df[str(length*1) + " MA"] > df[str(l*1) + " MA"]) & (df["Close"] > df[str(l*1) + " MA"])), f"Strategy {i} on {Time} Mins"] = "Buy"

Instead of this:

df.loc[( (df[str(length*Time) + " MA"] > df[str(l*Time) + " MA"]) & (df["Close"] > df[str(l*Time) + " MA"])), f"Strategy {i} on {Time} Mins"] = "Buy"