Picking a function to execute

[palladium]

I want to add moving averages of prices and volume of various periods of the DJIA index. The code below works:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

import pandas as pd   mydata = pd.read_csv('djia.csv', index_col = 'Date')   what_dict = {'price': ['price', 'Close'],                 'vol': ['vol', 'Volume']}   def add_avg(df = mydata, df_new, what):     """Adds moving averages of various rolling periods"""     """Params: df : original dataframe."""     """ df_new: new dataframe"""     """what: specifies either average price or volume"""       x = what_dict.get(what)       df_new[f'avg_{x[0]}_5'] = df[x[1]].rolling(5).mean().shift(1) #weekly avg     df_new[f'avg_{x[0]}_30'] = df[x[1]].rolling(21).mean().shift(1) #monthly avg     df_new[f'avg_{x[0]}_365'] = df[x[1]].rolling(252).mean().shift(1)   def add_std(df = mydata, df_new, what):     """Same as above, but adds std deviation instead"""       x = what_dict.get(what)       df_new[f'std_{x[0]}_5'] = df[x[1]].rolling(5).std().shift(1)     df_new[f'std_{x[0]}_30'] = df[x[1]].rolling(21).std().shift(1)     df_new[f'std_{x[0]}_365'] = df[x[1]].rolling(252).std().shift(1)   data = pd.DataFrame() list = ['price', 'vol'] for i in list:     add_col(data_raw, data, i)     add_std(data_raw, data, i)   print(data.columns)

Gives:

Output:
Index(['avg_price_5', 'avg_price_30', 'avg_price_365', 'std_price_5',
       'std_price_30', 'std_price_365', 'avg_vol_5', 'avg_vol_30',
       'avg_vol_365', 'std_vol_5', 'std_vol_30', 'std_vol_365', 'return_1',
       'return_5', 'return_30', 'return_365', 'moving_avg_5', 'moving_avg_30',
       'moving_avg_365'],
      dtype='object')

As you can see there is too much repetition. I want to combine the two functions into one. Specifically, I want to add an argument to the function to specify whether mean() or std() is called. Something like:

1 2 3 4 5 6 7 8 9

def add_col(df = mydata, df_new, what, func):    """func argument will specify whether mean() or std() to be called"""       func_dict = {'avg': 'mean', 'std': 'std'} #can't use mean() and std() as value here - throws an error saying they are     #not defined       y = func_dict.get(func)     df_new[f'avg_{x[0]}_5'] = df[x[1]].rolling(5).y().shift(1) #Throws Attribute Error - Rolling Object has no attribute 'y'     df_new[f'avg_{x[0]}_30'] = df[x[1]].rolling(21).y().shift(1)     df_new[f'avg_{x[0]}_365'] = df[x[1]].rolling(252).y().shift(1)

Would appreciate some pointers here.

Thank you.

Sample Data:

Output:
Date	    Open	    High	    Low	        Close	    Adj Close	Volume
1/12/2005	10806.03027	10934.90039	10806.03027	10912.57031	10912.57031	2569800
2/12/2005	10912.00977	10921.37012	10861.66016	10877.50977	10877.50977	2149000
5/12/2005	10876.9502	10876.9502	10810.66992	10835.00977	10835.00977	2373400
6/12/2005	10835.41016	10936.2002	10835.41016	10856.86035	10856.86035	2646300
7/12/2005	10856.86035	10868.05957	10764.00977	10810.91016	10810.91016	2434900
8/12/2005	10808.42969	10847.25	10729.66992	10755.12012	10755.12012	2532900
9/12/2005	10751.75977	10805.9502	10729.91016	10778.58008	10778.58008	2389300

[deanhystad]

You could use eval.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

import pandas as pd    mydata = pd.read_csv('Debug.txt', index_col = 'Date')    def add_col(src, dest, x, func):     dest[f'{func}_{x[0]}_5'] = eval(f'src[x].rolling(5).{func}().shift(1)')     dest[f'{func}_{x[0]}_30'] = eval(f'src[x].rolling(21).{func}().shift(1)')     dest[f'{func}_{x[0]}_365'] = eval(f'src[x].rolling(252).{func}().shift(1)')   data = pd.DataFrame() for column in ['Close', 'Volume']:     add_col(mydata, data, column, 'std')     add_col(mydata, data, column, 'mean')    print(data.columns)