I have a boxplot below (using seaborn) where the "box" part is too squashed. How do I change the scale along the y-axis so that the boxplot is more presentable (ie. the "box" part is too squashed) but still keeping all the outliers in the plot?
Many thanks.
You can do two things here.
Make the plot bigger
Change the range of the y-axis
Since you want to keep the outliers, rescaling the y-axis may not be that effective. You haven't given any data or code examples. So I'll just add a way to make your figure bigger.
# this script makes the figure bigger and rescale the y-axis
ax = plt.figure(figsize=(20,15))
ax = sns.boxplot(x="day", y="total_bill", data=tips)
ax.set_ylim(0,100)
You could set the axis after the plot:
import seaborn as sns
df = sns.load_dataset('iris')
a = sns.boxplot(y=df["sepal_length"])
a.set(ylim=(0,10))
Additionally, you could try dropping outliers from the plot passing showfliers = False in boxplot.
Related
I am practicing with Python Pandas plotting functions and I am trying to plot the content of two series extracted from the same dataframe into one plot.
When I plot the two series individually the result is correct. However, when I plot them together, the one that I plot as second appears flat in the picture.
Here is my code:
# dailyFlow and smooth are created in the same way from the same dataframe
dailyFlow = pd.Series(dataFrame...
smooth = pd.Series(dataFrame...
# lower the noise in the signal with standard deviation = 6
smooth = smooth.resample('D').sum().rolling(31, center=True, win_type='gaussian').sum(std=6)
dailyFlow.plot(style ='-b')
plt.legend(loc = 'upper right')
plt.show()
smooth.plot(style ='-r')
plt.legend(loc = 'upper right')
plt.show()
plt.figure(figsize=(12,5))
smooth.plot(style ='-r')
dailyFlow.plot(style ='-b')
plt.legend(loc = 'upper right')
plt.show()
Here is the output of my function:
I already tried using the parameter secondary_y=True in the second plot, but then I lose the information on the second line in the legend and the scaling between the two plots is wrong.
Many sources on the Internet seem to suggest that plotting the two series like I am doing should be correct, but then why is the third plot incorrect?
Thank you very much for your help.
For the data you have, the 3rd plot is correct. Look at the scale of the y axis on your two plots: one goes up to 70,000 and the other to 60,000,000.
I suspect what you actually want is a .rolling(...).mean() which should have a range comparable to your original data.
If you would like to make both plots bigger, you cold try something like this
fig, ax1 = plt.subplots()
ax1.set_ylim([0, 75000])
# plot first graph
ax2 = ax1.twinx() # second axes that shares the same x-axis
ax2.set_ylim([0, 60000000])
#plot the second graph
I am toying around with seaborn violinplot, trying to make a single "violin" with each half being a different distribution, to be easily compared.
Modifying the simple example from here by changing the x axis to x=smoker I got to the following graph (linked below).
import seaborn as sns
sns.set(style="whitegrid", palette="pastel", color_codes=True)
# Load the example tips dataset
tips = sns.load_dataset("tips")
# Draw a nested violinplot and split the violins for easier comparison
sns.violinplot(x="smoker", y="total_bill", hue="smoker",
split=True, inner="quart", data=tips)
sns.despine(left=True)
This is the resulting graph
I would like that the graph does not show two separated halves, just one single violin with two different distributions and colours.
Is it possible to do this with seaborn? Or maybe with other library?
Thanks!
This is because you are specifying two things for the x axis with this line x="smoker". Namely, that it plot smoker yes and smoker no.
What you really want to do is plot all data. To do this you can just specify a single value for the x axis.
sns.set(style="whitegrid", palette="pastel", color_codes=True)
# Load the example tips dataset
tips = sns.load_dataset("tips")
# Draw a nested violinplot and split the violins for easier comparison
sns.violinplot(x=['Data']*len(tips),y="total_bill", hue="smoker",
split=True, inner="quart",
palette={"Yes": "y", "No": "b"},
data=tips)
sns.despine(left=True)
This outputs the following:
I have just started using seaborn to produce my figures. However I can't seem to remove one of the legends produced here.
I am trying to plot two accuracies against each other and draw a line along the diagonal to make it easier to see which has performed better (if anyone has a better way of plotting this data in seaborn - let me know!). The legend I'd like to keep is the one on the left, that shows the different colours for 'N_bands' and different shapes for 'Subject No'
ax1 = sns.relplot(y='y',x='x',data=df,hue='N bands',legend='full',style='Subject No.',markers=['.','^','<','>','8','s','p','*','P','X','D','H','d']).set(ylim=(80,100),xlim=(80,100))
ax2 = sns.lineplot(x=range(80,110),y=range(80,110),legend='full')
I have tried setting the kwarg legend to 'full','brief' and False for both ax1 and ax2 (together and separately) and it only seems to remove the one on the left, or both.
I have also tried to remove the axes using matplotlib
ax1.ax.legend_.remove()
ax2.legend_.remove()
But this results in the same behaviour (left legend dissapearing).
UPDATE: Here is a minimal example you can run yourself:
test_data = np.array([[1.,2.,100.,9.],[2.,1.,100.,8.],[3.,4.,200.,7.]])
test_df = pd.DataFrame(columns=['x','y','p','q'], data=test_data)
sns.set_context("paper")
ax1=sns.relplot(y='y',x='x',data=test_df,hue='p',style='q',markers=['.','^','<','>','8'],legend='full').set(ylim=(0,4),xlim=(0,4))
ax2=sns.lineplot(x=range(0,5),y=range(0,5),legend='full')
Although this doesn't reproduce the error perfectly as the right legend is coloured (I have no idea how to reproduce this error then - does the way my dataframe was created make a difference?). But the essence of the problem remains - how do I remove the legend on the right but keep the one on the left?
You're plotting a lineplot in the (only) axes of a FacetGrid produced via relplot. That's quite unconventional, so strange things might happen.
One option to remove the legend of the FacetGrid but keeping the one from the lineplot would be
g._legend.remove()
Full code (where I also corrected for the confusing naming if grids and axes)
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
test_data = np.array([[1.,2.,100.,9.],[2.,1.,100.,8.],[3.,4.,200.,7.]])
test_df = pd.DataFrame(columns=['x','y','p','q'], data=test_data)
sns.set_context("paper")
g=sns.relplot(y='y',x='x',data=test_df,hue='p',style='q',markers=['.','^','<','>','8'], legend='full')
sns.lineplot(x=range(0,5),y=range(0,5),legend='full', ax=g.axes[0,0])
g._legend.remove()
plt.show()
Note that this is kind of a hack, and it might break in future seaborn versions.
The other option is to not use a FacetGrid here, but just plot a scatter and a line plot in one axes,
ax1 = sns.scatterplot(y='y',x='x',data=test_df,hue='p',style='q',
markers=['.','^','<','>','8'], legend='full')
sns.lineplot(x=range(0,5),y=range(0,5), legend='full', ax=ax1)
plt.show()
I am trying to create violinplots that shows confidence intervals for the mean. I thought an easy way to do this would be to plot a pointplot on top of the violinplot, but this is not working since they seem to be using different indices for the xaxis as in this example:
import matplotlib.pyplot as plt
import seaborn as sns
titanic = sns.load_dataset("titanic")
titanic.dropna(inplace=True)
fig, (ax1,ax2,ax3) = plt.subplots(1,3, sharey=True, figsize=(12,4))
#ax1
sns.pointplot("who", "age", data=titanic, join=False,n_boot=10, ax=ax1)
#ax2
sns.violinplot(titanic.age, groupby=titanic.who, ax=ax2)
#ax3
sns.pointplot("who", "age", data=titanic, join=False, n_boot=10, ax=ax3)
sns.violinplot(titanic.age, groupby=titanic.who, ax=ax3)
ax3.set_xlim([-0.5,4])
print(ax1.get_xticks(), ax2.get_xticks())
gives: [0 1 2] [1 2 3]
Why are these plots not assigning the same xtick numbers to the 'who'-variable and is there any way I can change this?
I also wonder if there is anyway I can change the marker for pointplot, because as you can see in the figure, the point is so big so that it covers the entire confidence interval. I would like just a horizontal line if possible.
I'm posting my final solution here. The reason I wanted to do this kind of plot to begin with, was to display information about the distribution shape, shift in means, and outliers in the same figure. With mwaskom's pointers and some other tweaks I finally got what I was looking for.
The left hand figure is there as a comparison with all data points plotted as lines and the right hand one is my final figure. The thick grey line in the middle of the violin is the bootstrapped 99% confidence interval of the mean, which is the white horizontal line, both from pointplot. The three dotted lines are the standard 25th, 50th and 75th percentile and the lines outside that are the caps of the whiskers of a boxplot I plotted on top of the violin plot. Individual data points are plotted as lines beyond this points since my data usually has a few extreme ones that I need to remove manually like the two points in the violin below.
For now, I am going to to continue making histograms and boxplots in addition to these enhanced violins, but I hope to find that all the information is accurately captured in the violinplot and that I can start and rely on it as my main initial data exploration plot. Here is the final code to produce the plots in case someone else finds them useful (or finds something that can be improved). Lots of tweaking to the boxplot.
import matplotlib as mpl
import matplotlib.pyplot as plt
import seaborn as sns
#change the linewidth which to get a thicker confidence interval line
mpl.rc("lines", linewidth=3)
df = sns.load_dataset("titanic")
df.dropna(inplace=True)
x = 'who'
y = 'age'
fig, (ax1,ax2) = plt.subplots(1,2, sharey=True, figsize=(12,6))
#Left hand plot
sns.violinplot(df[y], groupby=df[x], ax=ax1, inner='stick')
#Right hand plot
sns.violinplot(df[y], groupby=df[x], ax=ax2, positions=0)
sns.pointplot(df[x],df[y], join=False, ci=99, n_boot=1000, ax=ax2, color=[0.3,0.3,0.3], markers=' ')
df.boxplot(y, by=x, sym='_', ax=ax2, showbox=False, showmeans=True, whiskerprops={'linewidth':0},
medianprops={'linewidth':0}, flierprops={'markeredgecolor':'k', 'markeredgewidth':1},
meanprops={'marker':'_', 'color':'w', 'markersize':6, 'markeredgewidth':1.5},
capprops={'linewidth':1, 'color':[0.3,0.3,0.3]}, positions=[0,1,2])
#One could argue that this is not beautiful
labels = [item.get_text() + '\nn=' + str(df.groupby(x).size().loc[item.get_text()]) for item in ax2.get_xticklabels()]
ax2.set_xticklabels(labels)
#Clean up
fig.suptitle('')
ax2.set_title('')
fig.set_facecolor('w')
Edit: Added 'n='
violinplot takes a positions argument that you can use to put the violins somewhere else (they currently just inherit the default matplotlib boxplot positions).
pointplot takes a markers argument that you can use to change how the point estimate is rendered.
I use autofmt_xdate to plot long x-axis labels in a readable way. The problem is, when I want to combine different subplots, the x-axis labeling of the other subplots disappears, which I do not appreciate for the leftmost subplot in the figure below (two rows high). Is there a way to prevent autofmt_xdate from quenching the other x-axis labels? Or is there another way to rotate the labels? As you can see I experimented with xticks and "rotate" as well, but the results were not satisfying because the labels were rotated around their center, which resulted in messy labeling.
Script that produces plot below:
from matplotlib import pyplot as plt
from numpy import arange
import numpy
from matplotlib import rc
rc("figure",figsize=(15,10))
#rc('figure.subplot',bottom=0.1,hspace=0.1)
rc("legend",fontsize=16)
fig = plt.figure()
Test_Data = numpy.random.normal(size=20)
fig = plt.figure()
Dimension = (2,3)
plt.subplot2grid(Dimension, (0,0),rowspan=2)
plt.plot(Test_Data)
plt.subplot2grid(Dimension, (0,1),colspan=2)
for i,j in zip(Test_Data,arange(len(Test_Data))):
plt.bar(i,j)
plt.legend(arange(len(Test_Data)))
plt.subplot2grid(Dimension, (1,1),colspan=2)
xticks = [r"%s (%i)" % (a,b) for a,b in zip(Test_Data,Test_Data)]
plt.xticks(arange(len(Test_Data)),xticks)
fig.autofmt_xdate()
plt.ylabel(r'$Some Latex Formula/Divided by some Latex Formula$',fontsize=14)
plt.plot(Test_Data)
#plt.setp(plt.xticks()[1],rotation=30)
plt.tight_layout()
#plt.show()
This is actually a feature of the autofmt_xdate method. From the documentation of the autofmt_xdate method:
Date ticklabels often overlap, so it is useful to rotate them and right align them. Also, a common use case is a number of subplots with shared xaxes where the x-axis is date data. The ticklabels are often long, and it helps to rotate them on the bottom subplot and turn them off on other subplots, as well as turn off xlabels.
If you want to rotate the xticklabels of the bottom right subplot only, use
plt.setp(plt.xticks()[1], rotation=30, ha='right') # ha is the same as horizontalalignment
This rotates the ticklabels 30 degrees and right aligns them (same result as when using autofmt_xdate) for the bottom right subplot, leaving the two other subplots unchanged.