Data visualization in Data Science: theory and practice

Andrey Lukyanenko

Senior DS @ Careem

Content

  • What is data visualization and why is it useful?
  • Classics of data visualization
  • Common software for making the plots
  • Chart types
  • Skills for professional-level visualizations
  • What should and shouldn't you do
  • Data Visualization for Machine Learning

Why visualizations are important

Why visualizations are important

Source

Tufte

  • Maximize data ink-ratio
  • Minimize lie factor
  • Minimize chartjunk
  • Use proper scales and clear labeling

Python. Matplotlib

Python. Plotly

import plotly.graph_objects as go

import pandas as pd
from datetime import datetime

df = pd.read_csv('https://raw.githubusercontent.com/plotly/datasets/master/finance-charts-apple.csv')

fig = go.Figure(data=[go.Candlestick(x=df['Date'],
                open=df['AAPL.Open'],
                high=df['AAPL.High'],
                low=df['AAPL.Low'],
                close=df['AAPL.Close'])])

fig.show()

Python. Altair

import altair as alt
from vega_datasets import data

source = data.cars()

alt.Chart(source).mark_circle(size=60).encode(
    x='Horsepower',
    y='Miles_per_Gallon',
    color='Origin',
    tooltip=['Name', 'Origin', 'Horsepower', 'Miles_per_Gallon']
).interactive()

R

library(ggplot2)

ggplot(mpg, aes(displ, hwy, colour = class)) + 
  geom_point()

Tableau

D3.js

Skills

  • Data visualization
  • Graphical design, interface design, UX
  • Storytelling
  • Color: Hue, Value/Lightness, Saturation/Chroma, RGB vs CMYK
  • Accessibility
  • Style guide

Style guide

Improvements

Improvements

Improvements

Improvements

Improvements

Improvements

Improvements

Improvements

  • Exploratory vs explanatory, показать то, что важно
  • Pie, 3D

Good examples

https://www.kaggle.com/code/spitfire2nd/enthusiast-to-data-professional-what-changes?scriptVersionId=51106095&cellId=18

https://www.kaggle.com/code/spitfire2nd/enthusiast-to-data-professional-what-changes?scriptVersionId=51106095&cellId=21

import seaborn as sns
penguins = sns.load_dataset("penguins")
sns.pairplot(penguins, hue="species")
import shap
explainer = shap.TreeExplainer(model1, X_train)
shap_values = explainer.shap_values(X_train)

shap.summary_plot(shap_values, X_train)

References

References

References

References

References

Contacts

Data Visualization talk

By Andrey Lukyanenko

Data Visualization talk

  • 220

More from Andrey Lukyanenko