CODEWITHSUNDEEP
pythonseabornvisualization
Ranit
Ranit

Reputation: 97

Autoscaling in seaborn of axis values and legend

So, I was learning to build deep learning models, and during the visualization part I plotted a scatterplot where x and y axis were longitude and latitude, respectively, and the hue was equal to house prices. Although the price values were in float format, the legend is showing in different scaled value.

import matplotlib.pyplot as plt
import seaborn as sns
df = pd.read_csv("../DATA/kc_house_data.csv")
sns.histplot(df['price'])

enter image description here

non_top_1_perct = df.sort_values('price',ascending=False).iloc[216:]
plt.figure(figsize=(12,8))
sns.scatterplot(x= 'long', y= 'lat', data =non_top_1_perct,
           edgecolor = None, alpha = 0.2, palette  ='RdYlGn', hue='price')

enter image description here

Here, in the first picture, one can notice that x-axis scale is in formatted into scientific notation. Also, in the legend box, if I want to show 7,700,000 instead of 1.6, how I have to rescale both of them?

Data: kaggle house data

Upvotes: 0

Views: 801

Answers (1)

mosc9575
mosc9575

Reputation: 6347

Here is an example how you can modify your legend based on the values of existing legend entries.

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

df = pd.DataFrame({
    'lat': [4, 24, 31, 2, 3],
    'long': [3, 5, 5, 6, 7], 
    'price':[35e6, 54899767, 57890789, 62890798, 70897871]
})
ax = sns.scatterplot(y="lat", x="long", data=df, hue='price')

# modify legend entries
handles = ax.get_legend().legendHandles
ax.legend(handles, [str(round(float(v.get_label())/1e6,1))+'m' for v in handles] , loc='upper left')

# disable scientific notation on y axis
ax.ticklabel_format(style='plain', axis='y')

basic scatter

You have to adapt this to your task.

Upvotes: 1

Related Questions