[FIXED] How do I plot a barchart from the datasets mean value and not its default 0 value in seaborn?

Issue

I'm trying to plot a bar graph showing departure from the mean. I looked through seaborn's documentation but have not figured out how I can plot the bar graph where its y-axis starts at the mean of the dataset and not 0. Essentially, I'm looking at how I can replicate this graph, but using seaborn.

Example Dataset:

{'Year': {0: 2000, 1: 2001, 2: 2002, 3: 2003, 4: 2004, 5: 2005, 6: 2006, 7: 2007, 8: 2008, 9: 2009, 10: 2010, 11: 2011, 12: 2012, 13: 2013, 14: 2014, 15: 2015, 16: 2016, 17: 2017, 18: 2018, 19: 2019, 20: 2020}, 'Temperature (Degrees F)': {0: 64.87, 1: 64.89, 2: 65.55, 3: 64.11, 4: 61.63, 5: 63.83, 6: 63.46, 7: 65.02, 8: 63.3, 9: 62.14, 10: 64.86, 11: 64.98, 12: 66.25, 13: 65.27, 14: 63.49, 15: 64.77, 16: 64.49, 17: 65.65, 18: 64.8, 19: 64.74, 20: 65.28}}

Current Code:

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

df_temp = pd.read_csv(r"C:\Users\rov\Desktop\CO\temp.csv", usecols=['Year', 'Temperature (Degrees F)'])

sns.set(style='darkgrid')
sns.set_context('talk')

fig, (ax1) = plt.subplots(nrows=1, figsize=(15,5))
sns.barplot(x='Year', y='Temperature (Degrees F)', data=df_temp, color='#3792CB', edgecolor= 'black', ax=ax1)

Solution

Internally, seaborn.barplot calles matplotlib.axes.Axes.bar, and any keyword arguments not recognised by seaborn.barplot are passed through to the matplotlib bar method (see the seaborn barplot documentation for more information, specifically the **kwargs argument).

This second function has an argument bottom, which indicates the y coordinate of the base of the bar (see the matplotlib bar documentation for more information). You can set this bottom argument equal to the mean temperature value. However, the bar heights are determined relative to the bottom, so you should also subtract the mean temperature value from the bar heights. One example of how to achieve this is given below.

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

df_temp = pd.read_csv("temp.csv", usecols=['Year', 'Temperature (Degrees F)'])

# calculate the mean temperature and subtract it from the dataset
mean = df_temp['Temperature (Degrees F)'].mean()
df_temp['Temperature (Degrees F)'] -= mean

sns.set(style='darkgrid')
sns.set_context('talk')

fig, (ax1) = plt.subplots(nrows=1, figsize=(15,5))
sns.barplot(x='Year', y='Temperature (Degrees F)', data=df_temp,
            color='#3792CB', edgecolor= 'black', ax=ax1,
            bottom=mean)  # <-- set the bottom of the bars

Answered By - luuk