I have a dataframe with two columns. The power column represents the power consumption of the system. And the component_status column divide the data in two, based when the component is OFF or ON. When the values are 153 is when the component is ON and when the values are 150 the component is OFF.
The result that I am looking for is to have a boxplot with three boxplots, using sns.boxplot
. One is the power consumption with all the data, called "TOTAL". The other two, the power consumption based if the component was OFF or ON, called "COMPONENT = ON" "COMPONENT = OFF".
The data frame example is as follows:
power|component_status |
0.5 | 150 |
1.5 | 150 |
2.5 | 150 |
0.3 | 153 |
0.5 | 153 |
1.5 | 153 |
2.5 | 150 |
0.3 | 153 |
thanks for the help.
CodePudding user response:
Your first step is to build your data frame with the conditions. There are a few ways to go about this.
Let's start with an initial
df1
(dataframe #1) as you have given. Then, let's add acondition
column to say "Total". You can useprint(df1)
to see what this looks like.Then let's copy that dataframe into
df2
, and let's replace theconditions
with the off/on criteria from thecomponent_status
.Our final dataframe
df
is just a concatenation ofdf1
anddf2
.Now we have a dataframe
df
that is ready to go in Seaborn.### Set up import pandas as pd import numpy as np import seaborn as sns power = [0.5, 1.5, 2.5, 0.3, 0.5, 1.5, 2.5, 0.3] component_status = [150, 150, 150, 153, 153, 153, 150, 153] df1 = pd.DataFrame( data=zip(power, component_status), columns=["power", "component_status"] ) ### Step 1 df1["condition"] = "Total" # print(df1) ### Step 2 df2 = df1.copy() df2["condition"] = np.where(df2["component_status"] == 153, "On", "Off") ### If you have several criteria, it can be easier to use np.select ### ... or just use Pandas directly: # df2.loc[(df2['component_status'] == 153), 'condition'] = 'On' # df2.loc[(df2['component_status'] == 150), 'condition'] = 'Off' ### Step 3 df = pd.concat([df1,df2]) print(df) ### Step 4 sns.boxplot(data=df, x='condition', y='power')