通过 function 迭代行运行 df

Question

So I am trying execute a buy/sell action based on the 5 period moving average exceeding that of the 20 period moving average. After receiving too many errors I decided to do it separately for each relevant df instead of writing a function for it. An image of the final product can be seen below (last column of the screenshot is what I was hoping to achieve using one of these two functions). When I played around with these two functions I also came across the error which was 'key is invalid' which I didn't know how to solve. Any information would be much appreciated

def action(df):
    for i in range(df.index[]):
        if df.loc[i,'PointMovingAverage5'] > df.loc[i,'PointMovingAverage20'] & df.loc[i-1,'PointMovingAverage5'] < df.loc[i-1,'PointMovingAverage20']:
            df.loc[i, 'BUY/SELL'] = 'BUY'
        elif df.loc[i,'PointMovingAverage5'] < df.loc[i,'PointMovingAverage20'] & df.loc[i-1,'PointMovingAverage5'] > df.loc[i-1,'PointMovingAverage20']:
            df.loc[i, 'BUY/SELL'] = 'SELL'
        else:
            df.loc[i, 'BUY/SELL'] = 'NaN'
     
CBA['BUY/SELL'] = CBA.apply(action, axis = 1)

ERROR: 'str' object cannot be interpreted as an integer

def Action(df):
    for label, row in df.iterrows():
        if (df[row,'PointMovingAverage5'] > df[row,'PointMovingAverage20']) & (df[row-1,'PointMovingAverage5'] < df[row-1,'PointMovingAverage20']):
            df.loc[label, 'BUY/SELL'] = 'BUY'
        elif (df[row,'PointMovingAverage5'] < df[row,'PointMovingAverage20']) & (df[row-1,'PointMovingAverage5'] > df[row-1,'PointMovingAverage20']):
            df.loc[label, 'BUY/SELL'] = 'SELL'
        else:
            df.loc[label, 'BUY/SELL'] = 'NO ACTION'
        return df
           
CBA = CBA.apply(Action, axis = 1) 

ERROR: 'Series' object has no attribute 'iterrows'

End Goal

Answer 1

The reason it's not working is that iterrows returns (index, row), so the row value is the actual row of the dataframe and label is the column index. Also, the first index will be zero, so you can't subtract one from it and you need to account for that. I created the function below fixing those issues. Note that the function won't work if the averages go from negative to zero to positive or vice versa.

def Action(df):
    df = df.copy()
    df['BUY/SELL'] = 'NO ACTION'
    for i, row in df.iterrows():
        if i == 0:
            if (row['PointMovingAverage5'] < row['PointMovingAverage20']) & (df.loc[i+1,'PointMovingAverage5'] > df.loc[i+1,'PointMovingAverage20']):
                df.loc[i, 'BUY/SELL'] = 'BUY'
            elif (row['PointMovingAverage5'] > row['PointMovingAverage20']) & (df.loc[i+1,'PointMovingAverage5'] < df.loc[i+1,'PointMovingAverage20']):
                df.loc[i, 'BUY/SELL'] = 'SELL'
            continue
        if (row['PointMovingAverage5'] > row['PointMovingAverage20']) & (df.loc[i-1,'PointMovingAverage5'] < df.loc[i-1,'PointMovingAverage20']):
            df.loc[i, 'BUY/SELL'] = 'BUY'
        elif (row['PointMovingAverage5'] < row['PointMovingAverage20']) & (df.loc[i-1,'PointMovingAverage5'] > df.loc[i-1,'PointMovingAverage20']):
            df.loc[i, 'BUY/SELL'] = 'SELL'
    return df.copy()