Get started

package: seaborn
load the package

import seaborn as sns

Seaborn is a Python data visualization library based on matplotlib.
It provides a high-level interface for drawing attractive and informative statistical graphics.

Overview of the seaborn functionality

tips dataset

dataset for scatter plot

tips = sns.load_dataset("tips")
tips.head()

##    total_bill   tip     sex smoker  day    time  size
## 0       16.99  1.01  Female     No  Sun  Dinner     2
## 1       10.34  1.66    Male     No  Sun  Dinner     3
## 2       21.01  3.50    Male     No  Sun  Dinner     3
## 3       23.68  3.31    Male     No  Sun  Dinner     2
## 4       24.59  3.61  Female     No  Sun  Dinner     4

if you have trouble load the dataset, try the following code

import ssl
ssl._create_default_https_context = ssl._create_unverified_context

relplot (scatter plot)

scatter plot:
- data: a dataframe
- x: x axis
- y: y axis
implementation:
- sns.relplot with kind: scatter (default)
- sns.scatterplot

sns.relplot(data=tips,x="total_bill", y="tip", kind='scatter')

# sns.scatterplot(data=tips, x="total_bill", y="tip", )

color and style

hue: Grouping variable that will produce elements with different colors

sns.relplot(data=tips,x="total_bill", y="tip", hue = "smoker")

style: shape of the points

sns.relplot(data=tips,x="total_bill", y="tip", hue = "smoker", style="smoker")

size and transparency

size: size of points

sns.relplot(data=tips,x="total_bill", y="tip", size = "size")

alpha: transparency

sns.relplot(data=tips,x="total_bill", y="tip", size = "size", alpha = 0.5)

multiple subfigures

col: columns

sns.relplot(
    data=tips,
    x="total_bill", y="tip", col="time",
    hue="smoker", style="smoker", size="size",
)

scatter plot with linear regression fitting

sns.lmplot(data=tips, x="total_bill", y="tip", col="time", hue="smoker")

styles

There are 5 built-in styles
- darkgrid (default)
- whitegrid
- dark
- white
- ticks

sns.set_style("whitegrid")
sns.relplot(data=tips,x="total_bill", y="tip", kind='scatter')

alternative function:

sns.set_theme(style = "whitegrid")

Scaling Plots – set_context

There are 4 built-in context (related to font size)
- notebook (default)
- paper
- talk
- poster

sns.set_style("whitegrid")
sns.set_context("talk")
sns.relplot(data=tips,x="total_bill", y="tip", kind='scatter')

alternative function:

sns.set_theme(context = "talk, style = "whitegrid")

fmri dataset

dataset for line plot

fmri = sns.load_dataset("fmri")
fmri.head()

##   subject  timepoint event    region    signal
## 0     s13         18  stim  parietal -0.017552
## 1      s5         14  stim  parietal -0.080883
## 2     s12         18  stim  parietal -0.081033
## 3     s11         18  stim  parietal -0.046134
## 4     s10         18  stim  parietal -0.037970

fmri_sub13 = fmri[(fmri["subject"]=="s13")]
fmri_sub13_stim = fmri_sub13[(fmri_sub13["event"]=="stim")]
fmri_sub13_stim_parietal = fmri_sub13_stim[(fmri_sub13_stim["region"]=="parietal")]
fmri_sub13_stim_parietal.head()

##    subject  timepoint event    region    signal
## 0      s13         18  stim  parietal -0.017552
## 15     s13         17  stim  parietal -0.008265
## 29     s13         16  stim  parietal -0.002856
## 43     s13         15  stim  parietal -0.010971
## 57     s13         14  stim  parietal -0.033713

relplot (line plot)

line plot:
- data: a dataframe
- x: x axis
- y: y axis
implementation:
- sns.relplot with kind = line
- sns.lineplot

sns.relplot(data=fmri_sub13_stim_parietal, kind="line", x="timepoint", y="signal")

# sns.lineplot(data=fmri_sub13_stim_parietal, x="timepoint", y="signal")

combine scatter and line plots

fig, ax = plt.subplots(figsize=(6, 4))
sns.scatterplot(data=fmri_sub13_stim_parietal, x="timepoint", y="signal", ax=ax)
sns.lineplot(data=fmri_sub13_stim_parietal, x="timepoint", y="signal", ax=ax)

relplot (line plot)

line plot:
- hue: Grouping variable that will produce elements with different colors
- style: shape
- size: size
- col: columns
color by hue

sns.relplot(data=fmri_sub13_stim, kind="line", x="timepoint", y="signal", hue="region", )

style

sns.relplot(data=fmri_sub13, kind="line", x="timepoint", y="signal", hue="region", style = "event", )

columns

sns.relplot(data=fmri_sub13, kind="line", x="timepoint", y="signal", hue="region", col="event", )

multiple subjects

If multiple data exist for the same data points, the mean value and the confidence interval will be plotted

sns.relplot(data=fmri, kind="line", x="timepoint", y="signal", hue="region", col="event", )

Exercise

based on the following fmri sub dataset, draw a spaghetti plot of signal with respect to time point for each subject

fmri_sub = fmri[(fmri["event"]=="stim") & (fmri["region"]=="parietal")]
fmri_sub.head()

##   subject  timepoint event    region    signal
## 0     s13         18  stim  parietal -0.017552
## 1      s5         14  stim  parietal -0.080883
## 2     s12         18  stim  parietal -0.081033
## 3     s11         18  stim  parietal -0.046134
## 4     s10         18  stim  parietal -0.037970

solution

sns.relplot(data=fmri_sub, kind="line", x="timepoint", y="signal", hue="subject", )

histogram

sns.displot
- kind=“hist” by default
- kind=“kde”: kernal density estimation
- kind=“ecdf”: emperical cdf

# sns.displot(data=tips, x="total_bill")
sns.displot(data=tips, x="total_bill", col="time", kde=True)

## density function only 
## sns.displot(data=tips, x="total_bill", col="time", kind = "kde")

empirical cumulative distribution function

sns.displot(data=tips, kind="ecdf", x="total_bill", col="time", hue="smoker")

rug: show each observation with marginal ticks

# sns.displot(data=tips, kind="ecdf", x="total_bill", col="time", hue="smoker", rug=True)

Visualizing categorical data

Categorical scatterplots:
- stripplot() (with kind=“strip”; the default)
- swarmplot() (with kind=“swarm”)
Categorical distribution plots:
- boxplot() (with kind=“box”)
- violinplot() (with kind=“violin”)
Categorical estimate plots:
- pointplot() (with kind=“point”)
- barplot() (with kind=“bar”)
- countplot() (with kind=“count”)

catplot (strip plot)

strip plot: jitter plot

sns.catplot(data=tips, kind="strip", x="day", y="total_bill", hue="smoker")

sns.catplot(data=tips, kind="strip", x="day", y="total_bill", hue="smoker", dodge=True)
sns.stripplot(data=tips, x="day", y="total_bill", hue="smoker")

catplot (swarm plot)

swarm plot

sns.catplot(data=tips, kind="swarm", x="day", y="total_bill", hue="smoker", dodge=True)

sns.catplot(data=tips, kind="swarm", x="day", y="total_bill", hue="smoker", )
sns.swarmplot(data=tips, x="day", y="total_bill", hue="smoker")

catplot (boxplot)

boxplot
- sns.catplot with kind = “box”
- sns.boxplot

sns.catplot(data=tips, kind="box", x="day", y="total_bill", hue="smoker")

# sns.boxplot(data=tips, x="day", y="total_bill", hue="smoker")

overlay box plot and jitter plot

fig, ax = plt.subplots(figsize=(6, 4))
sns.stripplot(data=tips, x="day", y="total_bill", hue="smoker", dodge=True, ax=ax, legend=False)
sns.boxplot(data=tips, x="day", y="total_bill", hue="smoker", ax=ax)

catplot (violin)

violin
- sns.catplot with kind = “violin”
- sns.violinplot

sns.catplot(data=tips, kind="violin", x="day", y="total_bill", hue="smoker")

# sns.violinplot(data=tips, x="day", y="total_bill", hue="smoker")

catplot (pointplot)

pointplot: point estimate + error bar
- sns.catplot with kind = “point”
- sns.pointplot

sns.catplot(data=tips, kind="point", x="day", y="total_bill", hue="smoker")

# sns.pointplot(data=tips, x="day", y="total_bill", hue="smoker")

catplot (barplot)

barplot:
- sns.catplot with kind = “bar”
- sns.barplot

sns.catplot(data=tips, kind="bar", x="day", y="total_bill", hue="smoker")

# sns.barplot(data=tips, x="day", y="total_bill", hue="smoker")
# sns.barplot(data=tips, x="day", y="total_bill", hue="smoker", ci=None)
# sns.barplot(data=tips, y="day", x="total_bill", hue="smoker") ## horizontal barplot

catplot (countplot)

countplot
- sns.catplot with kind = “count”
- sns.countplot

sns.catplot(data=tips, kind="count", x="day", hue="smoker")

# sns.countplot(data=tips, x="day", hue="smoker")

iris dataset

dataset for heatmap and pairplot

iris = sns.load_dataset("iris")
iris.head()

##    sepal_length  sepal_width  petal_length  petal_width species
## 0           5.1          3.5           1.4          0.2  setosa
## 1           4.9          3.0           1.4          0.2  setosa
## 2           4.7          3.2           1.3          0.2  setosa
## 3           4.6          3.1           1.5          0.2  setosa
## 4           5.0          3.6           1.4          0.2  setosa

pairplot

to visualize the relationship between each pair of numerical variable

p1 = sns.pairplot(iris, hue="species")
#sns.pairplot(iris)
plt.show()

heatmap

iris_sub = iris.drop(columns="species")
sns.heatmap(iris_sub)

Programming basics for Biostatistics 6099

seaborn: statistical data visualization

Outlines

Get started

Overview of the seaborn functionality

tips dataset

relplot (scatter plot)

color and style

size and transparency

multiple subfigures

scatter plot with linear regression fitting

styles

Scaling Plots – set_context

fmri dataset

relplot (line plot)

combine scatter and line plots

relplot (line plot)

multiple subjects

Exercise

histogram

empirical cumulative distribution function

Visualizing categorical data

catplot (strip plot)

catplot (swarm plot)

catplot (boxplot)

overlay box plot and jitter plot

catplot (violin)

catplot (pointplot)

catplot (barplot)

catplot (countplot)

iris dataset

pairplot

heatmap

Reference