[SOLVED] How to get the N'th row of average of max,min of two forex columns in pandas

How to get the N'th row of average of max,min of two forex columns in pandas

Here's a little part of my data frame of forex data:

OPEN    HIGH    LOW     CLOSE   VOLUME
1,14257 1,14368 1,14234 1,14325 1474
1,14327 1,14358 1,14302 1,14334 1032
1,14339 1,14401 1,14232 1,14391 993
1,14393 1,14431 1,14371 1,14376 1401
1,14377 1,14416 1,14342 1,14414 812
1,14419 1,14474 1,14393 1,14428 1091
1,14426 1,14426 1,14375 1,14405 1190
1,14405 1,14412 1,14366 1,14388 991
1,14388 1,14439 1,14384 1,14389 781

I want to implement the following calculation on a pandas dataframe.

Example: Middle value between 'OPEN' and 'CLOSE' highest and lowest value within the 9 last candles.

I know how to do it in Excel, the 10th line will get the following value in a new column: =(MAX(B2:B10;E2:E10)+MIN(B2:B10;E2:E10))/2
I also have some idea to do it in a 2D table using two ints in two while-loops as the table indexes, but it doesn't seems to work with a DataFrame.

Context:

I'm learning pandas and pandas_ta packages and following some YouTube videos. I want to create a forex trading bot that gets the values from a Forex API and sends me an email when it's time to buy or sell.

I found someone using Machine Learning for a forex bot, creating a pandas dataframe from a CSV file to teach his "IA" and adding some values in it using pandas_ta.

Solution

many TA signals can be calculated using pandas rolling function: https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.rolling.html https://pandas.pydata.org/docs/reference/api/pandas.Series.rolling.html

for this specific case, try this:

max_signal = df[['OPEN', 'CLOSE']].max(axis=1)
min_signal = df[['OPEN', 'CLOSE']].min(axis=1)
df['signal'] = (max_signal.rolling(9).max() + min_signal.rolling(9).min()) / 2

use fillna if you don't want the NaN values