'Pandas: groupyby with 'nth' sorting even though sort assigned to FALSE
GIVEN
A database df (like below).
date time open high low close
18-10-2010 9:15 6066 6079 6050 6052
18-10-2010 9:25 6050 6053 6034 6053
.
.
18-10-2010 15:25 6123 6136 6123 6131.5
19-10-2010 9:15 6108 6109 6094 6096.75
.
.
03-03-2022 15:25 16524 16528 16518 16528
MY GOAL
I want to extract the highest high, lowest low and close of the day for everyday. I have the following code for that.
day_HLC = df.groupby('day', sort=False).high.max()
day_HLC = day_HLC.reset_index()
day_HLC['low'] = df.groupby('day', sort=False).low.min().reset_index().low
test = df.groupby('day', sort=False).close.nth(-1).reset_index()
MY PROBLEM
The highest high and the lowest low are working out perfectly in the right order, without any sorting, as required. But the code for close of the day is giving me a sorted pandas series that is creating problems for further calculations.
highest high and lowest low working as I want:
day high low
0 18-10-2010 6136.00 6007.60
1 19-10-2010 6166.40 6024.10
test failing and giving me a sorted result
day close
0 01-01-2013 6007.00
1 01-01-2014 6360.65
2 01-01-2015 8347.05
3 01-01-2016 7983.00
4 01-01-2018 10490.00
QUESTION
How can I get a non sorted result for the close of everyday without it being sorted?
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|