About cumprod and cummax

1. Purpose

Here is an example of how to use cumprod and cummax.

2 Contents

cumprod ・ ・ ・ Calculates the cumulative product for all elements of the array. cummax ・ ・ ・ Extracts the maximum value from all the elements of the array.

test.py


import numpy as np
import pandas as pd


dat = [
    ['2019-07-01',10],
    ['2019-07-02',12],
    ['2019-07-03',8],
    ['2019-07-04',14],
    ['2019-07-05',7],
    ['2019-07-08',3]
]
df0 = pd.DataFrame(dat,columns=["A","B"])

print(df0)

df1=df0['B'].pct_change()
df2=(1 + df1)
df3=(1 + df1).cumprod() #Calculates the cumulative product from the 0th to the i-th element.
df4=(1 + df1).cummax()  #Extracts the maximum value among the 0th to i-th elements.
print(df1)
print(df2)
print(df3)
print(df4)
<df0:Basic data>
            A   B
0  2019-07-01  10
1  2019-07-02  12
2  2019-07-03   8
3  2019-07-04  14
4  2019-07-05   7
5  2019-07-08   3

df1: (df0['B']I th and i-Calculate the first rate of change)
0         NaN
1    0.200000
2   -0.333333
3    0.750000
4   -0.500000
5   -0.571429
Name: B, dtype: float64

df2: (df1+Calculate 1)
0         NaN
1    1.200000
2    0.666667
3    1.750000
4    0.500000
5    0.428571
Name: B, dtype: float64

df2: (cumprod()Calculate the cumulative product)

0    NaN
1    1.2
2    0.8  (=1.2×0.666667)
3    1.4   (=1.2×0.666667×1.75)
4    0.7   (=1.2×0.666667×1.75×0.5)
5    0.3   (=1.2×0.666667×1.75×0.5×0.428)
Name: B, dtype: float64

df3: (cummax()Extract the maximum value from the elements.)

0     NaN
1    1.20   (df2:Maximum value among elements 1 to 1)
2    1.20   (df2:Maximum value among elements 1 to 2)
3    1.75   (df2:Maximum value among elements 1 to 3)
4    1.75   (df2:Maximum value among elements 1 to 4)
5    1.75   (df2:Maximum value among elements 1 to 5)
Name: B, dtype: float64

Recommended Posts

About cumprod and cummax
About _ and __
About cross-validation and F-number
This and that about pd.DataFrame
Linux (about files and directories)
About python objects and classes
About Python variables and objects
About LINUX files and processes
About Raid group and LUN
About fork () function and execve () function
About Django's deconstruct and deconstructible
About Python, len () and randint ()
About Python datetime and timezone
About Sharpe Ratio and Sortino Ratio
About Python and regular expressions
About Python and os operations
About http.Handle () and http.NewServeMux (). Handle ()
Python # About reference and copy
About Numpy array and asarray
About Python sort () and reverse ()
About installing Pwntools and Python2 series
Summary and common errors about cron
About python dict and sorted functions
About dtypes in Python and Cython
About MkDocs themes and their customs
About Python pickle (cPickle) and marshal
[Python] About Executor and Future classes
About Python, from and import, as
About time series data and overfitting
About package management with conda and pip
Memorandum about regression and binary classification metrics
About sensor_mode and angle of view of picamera
About errors during PyInstaller installation and execution
About PyQt signal, connect and lambda expressions
Personal notes and links about machine learning ① (Machine learning)
[SSH] About public key authentication and EC2
Finding out about file permissions and superuser
About the relationship between Git and GitHub
A story about Python pop and append
About AntiDDoS Premium and Special Features TOA
Talking about old and new Python classes
[Golang] About Go language Producer and Consumer
Talking about Python class attributes and metaclasses