Related
This is probably a misunderstanding how colormaps are different from palettes, but I'd like to use a colormap that is not available in seaborn for coloring my binned dataset. I tried using palettable and now cmocean in particular directly but will get a TypeError;
'LinearSegmentedColormap' object is not iterable
Using any of the palettes that are available in Seaborn will work just fine, but I need a palette that doesn't go to white as this adds a weird 'banding' to the plot.
I have a dataframe with 3 columns with numerical data, dimensions and added a bin column for the colors usage in the plot.
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import cmocean
cmap=cmocean.cm.balance
cpal=sns.color_palette(cmap,n_colors=64,desat=0.2)
plt.style.use("seaborn-dark")
ax = sns.stripplot(x='Data', y='Dimension', data=dfBalance, jitter=0.15, edgecolor='none', alpha=0.4, size=4, hue='bin', palette=cpal)
sns.despine()
ax.legend_.remove()
plt.show()
Seaborn does not take a Colormap instance as input for .color_palette. It takes
name of matplotlib cmap, [...], or a list of colors in any format matplotlib accepts
Since cmocean registers its colormaps with matplotlib with a "cmo." prefix, you would do
import seaborn as sns
import cmocean
cpal = sns.color_palette("cmo.balance", n_colors=64, desat=0.2)
In case you have a custom colormap created yourself or from any other package, you might register it yourself.
import seaborn as sns
import matplotlib.cm
import matplotlib.colors
cmap = matplotlib.colors.LinearSegmentedColormap.from_list("", ["brown", "pink", "limegreen"])
matplotlib.cm.register_cmap("mycolormap", cmap)
cpal = sns.color_palette("mycolormap", n_colors=64, desat=0.2)
Is there a way to set the color bar scale to log on a seaborn heat map graph?
I am using a pivot table output from pandas as an input to the call
sns.heatmap(df_pivot_mirror,annot=False,xticklabels=256,yticklabels=128,cmap=plt.cm.YlOrRd_r)
Thank you.
If you have a current install of seaborn, norm=LogNorm() in the call to heatmap works now. (Pointed out in the comments -- thank you.) Adding this to one of the seaborn examples:
import numpy as np
import seaborn as sns; sns.set_theme(style='white')
import matplotlib.pyplot as plt
from matplotlib.colors import LogNorm, Normalize
from matplotlib.ticker import MaxNLocator
flights = sns.load_dataset("flights")
flights = flights.pivot("month", "year", "passengers")
f3, ax5 = plt.subplots(1,1)
sns.heatmap(flights, square=True, norm=LogNorm())
You can pass through colorbar arguments as keywords in the seaborn wrapper, but they sometimes collide with the seaborn choices:
sns.heatmap(flights, square=True, norm=LogNorm(), cbar_kws={'ticks':MaxNLocator(2), 'format':'%.e'})
For comparison, this is the matplotlib heatmap without seaborn's improvements -- the colorbar arguments have both been applied:
f5, ax6 = plt.subplots(1,1)
im6 = plt.imshow(flights, norm=LogNorm())
cbar6 = ax.figure.colorbar(im6, ax=ax6, ticks=MaxNLocator(2), format='%.e')
If you have to use an older install and LogNorm doesn't work in seaborn, see the previous versions of this answer for a workaround.
Short Answer:
from matplotlib.colors import LogNorm
sns.heatmap(df, norm=LogNorm())
You can normalize the values on the colorbar with matplotlib.colors.LogNorm.
I also had to manually set the labels in seaborn and ended up with the following code:
#!/usr/bin/env python3
import math
import numpy as np
import seaborn as sn
from matplotlib.colors import LogNorm
data = np.random.rand(20, 20)
log_norm = LogNorm(vmin=data.min().min(), vmax=data.max().max())
cbar_ticks = [math.pow(10, i) for i in range(math.floor(math.log10(data.min().min())), 1+math.ceil(math.log10(data.max().max())))]
sn.heatmap(
data,
norm=log_norm,
cbar_kws={"ticks": cbar_ticks}
)
Responding to cphlewis (I don't have enough reputation), I solved this problem using cbar_kws; as I saw here: seaborn clustermap: set colorbar ticks.
For example cbar_kws={"ticks":[0,1,10,1e2,1e3,1e4,1e5]}.
from matplotlib.colors import LogNorm
s=np.random.rand(20,20)
sns.heatmap(s, norm=LogNorm(s.min(),s.max()),
cbar_kws={"ticks":[0,1,10,1e2,1e3,1e4,1e5]},
vmin = 0.001, vmax=10000)
plt.show()
Have a nice day.
I'm working on the Iris data and trying to use scatter plot, while I was able to get the output, I'd like to know how I can color the points based on their species, using matplotlib.
I've using the following syntax:
iris.plot.scatter(x='petal_length', y='petal_width')
iris.plot(kind='scatter', x='sepal_length', y='sepal_width')
Also is there any way to use a single line of code to create two scatter plots for sepal_length/width and petal_length/width while coloring based on species?
Getting the colors correct in a single call to the plotting function is a bit tedious.
import seaborn as sns
iris = sns.load_dataset("iris")
import numpy as np
import matplotlib.pyplot as plt
u, inv = np.unique(iris.species.values, return_inverse=True)
ax = iris.plot.scatter(x='petal_length', y='petal_width',
c=inv, cmap="brg", colorbar=False)
plt.show()
I would hence recommend to loop over the species, with the additional advantage of being able to easily put a legend into the plot.
import seaborn as sns
iris = sns.load_dataset("iris")
import matplotlib.pyplot as plt
for n, grp in iris.groupby("species"):
plt.scatter(grp.petal_length, grp.petal_width, label=n)
plt.legend()
plt.show()
An easy solution is also to use seaborn.
import seaborn as sns
iris = sns.load_dataset("iris")
import matplotlib.pyplot as plt
g = sns.FacetGrid(iris, hue="species")
g.map(plt.scatter, 'petal_length','petal_width').add_legend()
plt.show()
Is there a way to set the color bar scale to log on a seaborn heat map graph?
I am using a pivot table output from pandas as an input to the call
sns.heatmap(df_pivot_mirror,annot=False,xticklabels=256,yticklabels=128,cmap=plt.cm.YlOrRd_r)
Thank you.
If you have a current install of seaborn, norm=LogNorm() in the call to heatmap works now. (Pointed out in the comments -- thank you.) Adding this to one of the seaborn examples:
import numpy as np
import seaborn as sns; sns.set_theme(style='white')
import matplotlib.pyplot as plt
from matplotlib.colors import LogNorm, Normalize
from matplotlib.ticker import MaxNLocator
flights = sns.load_dataset("flights")
flights = flights.pivot("month", "year", "passengers")
f3, ax5 = plt.subplots(1,1)
sns.heatmap(flights, square=True, norm=LogNorm())
You can pass through colorbar arguments as keywords in the seaborn wrapper, but they sometimes collide with the seaborn choices:
sns.heatmap(flights, square=True, norm=LogNorm(), cbar_kws={'ticks':MaxNLocator(2), 'format':'%.e'})
For comparison, this is the matplotlib heatmap without seaborn's improvements -- the colorbar arguments have both been applied:
f5, ax6 = plt.subplots(1,1)
im6 = plt.imshow(flights, norm=LogNorm())
cbar6 = ax.figure.colorbar(im6, ax=ax6, ticks=MaxNLocator(2), format='%.e')
If you have to use an older install and LogNorm doesn't work in seaborn, see the previous versions of this answer for a workaround.
Short Answer:
from matplotlib.colors import LogNorm
sns.heatmap(df, norm=LogNorm())
You can normalize the values on the colorbar with matplotlib.colors.LogNorm.
I also had to manually set the labels in seaborn and ended up with the following code:
#!/usr/bin/env python3
import math
import numpy as np
import seaborn as sn
from matplotlib.colors import LogNorm
data = np.random.rand(20, 20)
log_norm = LogNorm(vmin=data.min().min(), vmax=data.max().max())
cbar_ticks = [math.pow(10, i) for i in range(math.floor(math.log10(data.min().min())), 1+math.ceil(math.log10(data.max().max())))]
sn.heatmap(
data,
norm=log_norm,
cbar_kws={"ticks": cbar_ticks}
)
Responding to cphlewis (I don't have enough reputation), I solved this problem using cbar_kws; as I saw here: seaborn clustermap: set colorbar ticks.
For example cbar_kws={"ticks":[0,1,10,1e2,1e3,1e4,1e5]}.
from matplotlib.colors import LogNorm
s=np.random.rand(20,20)
sns.heatmap(s, norm=LogNorm(s.min(),s.max()),
cbar_kws={"ticks":[0,1,10,1e2,1e3,1e4,1e5]},
vmin = 0.001, vmax=10000)
plt.show()
Have a nice day.
How do I change the size of my image so it's suitable for printing?
For example, I'd like to use to A4 paper, whose dimensions are 11.7 inches by 8.27 inches in landscape orientation.
You can also set figure size by passing dictionary to rc parameter with key 'figure.figsize' in seaborn set method:
import seaborn as sns
sns.set(rc={'figure.figsize':(11.7,8.27)})
Other alternative may be to use figure.figsize of rcParams to set figure size as below:
from matplotlib import rcParams
# figure size in inches
rcParams['figure.figsize'] = 11.7,8.27
More details can be found in matplotlib documentation
You need to create the matplotlib Figure and Axes objects ahead of time, specifying how big the figure is:
from matplotlib import pyplot
import seaborn
import mylib
a4_dims = (11.7, 8.27)
df = mylib.load_data()
fig, ax = pyplot.subplots(figsize=a4_dims)
seaborn.violinplot(ax=ax, data=df, **violin_options)
Note that if you are trying to pass to a "figure level" method in seaborn (for example lmplot, catplot / factorplot, jointplot) you can and should specify this within the arguments using height and aspect.
sns.catplot(data=df, x='xvar', y='yvar',
hue='hue_bar', height=8.27, aspect=11.7/8.27)
See https://github.com/mwaskom/seaborn/issues/488 and Plotting with seaborn using the matplotlib object-oriented interface for more details on the fact that figure level methods do not obey axes specifications.
first import matplotlib and use it to set the size of the figure
from matplotlib import pyplot as plt
import seaborn as sns
plt.figure(figsize=(15,8))
ax = sns.barplot(x="Word", y="Frequency", data=boxdata)
You can set the context to be poster or manually set fig_size.
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
np.random.seed(0)
n, p = 40, 8
d = np.random.normal(0, 2, (n, p))
d += np.log(np.arange(1, p + 1)) * -5 + 10
# plot
sns.set_style('ticks')
fig, ax = plt.subplots()
# the size of A4 paper
fig.set_size_inches(11.7, 8.27)
sns.violinplot(data=d, inner="points", ax=ax)
sns.despine()
fig.savefig('example.png')
This can be done using:
plt.figure(figsize=(15,8))
sns.kdeplot(data,shade=True)
In addition to elz answer regarding "figure level" methods that return multi-plot grid objects it is possible to set the figure height and width explicitly (that is without using aspect ratio) using the following approach:
import seaborn as sns
g = sns.catplot(data=df, x='xvar', y='yvar', hue='hue_bar')
g.fig.set_figwidth(8.27)
g.fig.set_figheight(11.7)
This shall also work.
from matplotlib import pyplot as plt
import seaborn as sns
plt.figure(figsize=(15,16))
sns.countplot(data=yourdata, ...)
For my plot (a sns factorplot) the proposed answer didn't works fine.
Thus I use
plt.gcf().set_size_inches(11.7, 8.27)
Just after the plot with seaborn (so no need to pass an ax to seaborn or to change the rc settings).
See How to change the image size for seaborn.objects for a solution with the new seaborn.objects interface from seaborn v0.12, which is not the same as seaborn axes-level or figure-level plots.
Adjusting the size of the plot depends if the plot is a figure-level plot like seaborn.displot, or an axes-level plot like seaborn.histplot. This answer applies to any figure or axes level plots.
See the the seaborn API reference
seaborn is a high-level API for matplotlib, so seaborn works with matplotlib methods
Tested in python 3.8.12, matplotlib 3.4.3, seaborn 0.11.2
Imports and Data
import seaborn as sns
import matplotlib.pyplot as plt
# load data
df = sns.load_dataset('penguins')
sns.displot
The size of a figure-level plot can be adjusted with the height and/or aspect parameters
Additionally, the dpi of the figure can be set by accessing the fig object and using .set_dpi()
p = sns.displot(data=df, x='flipper_length_mm', stat='density', height=4, aspect=1.5)
p.fig.set_dpi(100)
Without p.fig.set_dpi(100)
With p.fig.set_dpi(100)
sns.histplot
The size of an axes-level plot can be adjusted with figsize and/or dpi
# create figure and axes
fig, ax = plt.subplots(figsize=(6, 5), dpi=100)
# plot to the existing fig, by using ax=ax
p = sns.histplot(data=df, x='flipper_length_mm', stat='density', ax=ax)
Without dpi=100
With dpi=100
# Sets the figure size temporarily but has to be set again the next plot
plt.figure(figsize=(18,18))
sns.barplot(x=housing.ocean_proximity, y=housing.median_house_value)
plt.show()
Some tried out ways:
import seaborn as sns
import matplotlib.pyplot as plt
ax, fig = plt.subplots(figsize=[15,7])
sns.boxplot(x="feature1", y="feature2",data=df) # where df would be your dataframe
or
import seaborn as sns
import matplotlib.pyplot as plt
plt.figure(figsize=[15,7])
sns.boxplot(x="feature1", y="feature2",data=df) # where df would be your dataframe
The top answers by Paul H and J. Li do not work for all types of seaborn figures. For the FacetGrid type (for instance sns.lmplot()), use the size and aspect parameter.
Size changes both the height and width, maintaining the aspect ratio.
Aspect only changes the width, keeping the height constant.
You can always get your desired size by playing with these two parameters.
Credit: https://stackoverflow.com/a/28765059/3901029