The following code:
# in ipython notebook, enable inline plotting with:
# %pylab inline --no-import-all
import matplotlib.pyplot as plt
# create some circles
circle1 = plt.Circle((-.5,0), 1, color='r', alpha=.2)
circle2 = plt.Circle(( .5,0), 1, color='b', alpha=.2)
# add them to the plot (bad form to use ;, but saving space)
# and control the display a bit
ax = plt.gca()
ax.add_artist(circle1); ax.add_artist(circle2)
ax.set_xlim(-2, 2); ax.set_ylim(-2, 2)
ax.set_aspect('equal')
# display it
plt.plot()
Produces the following plot:
I would like to specify the colors of the four regions (1) the background (currently white), (2 and 3) each individual event (the non-overlapping areas, currently blue and red), and (4) the intersection event (currently blended to purple). For example, I might color them red, green, blue, yellow -or- I might give them four different, precisely specified grayscale values (the later is more likely). [The colors will be generated based on characteristics of the underlying data.]
I specifically do not want to use alpha blending to "infer" a color in the intersection. I need to explicitly control the colors of all four regions.
I can think of a few strategies to solve this:
Ask mpl to extract the "primitive" patch objects that make up the three distinctly colored graphical regions (and do something similar to operate on the background) and then color them.
Given the circles, manually compute their intersections and color that intersection (somehow). Going point by point seems ugly.
Thanks!
I'm not 100% sure but I think matplotlib does not have the functionality to intersect polygons. But you could use shapely:
import shapely.geometry as sg
import matplotlib.pyplot as plt
import descartes
# create the circles with shapely
a = sg.Point(-.5,0).buffer(1.)
b = sg.Point(0.5,0).buffer(1.)
# compute the 3 parts
left = a.difference(b)
right = b.difference(a)
middle = a.intersection(b)
# use descartes to create the matplotlib patches
ax = plt.gca()
ax.add_patch(descartes.PolygonPatch(left, fc='b', ec='k', alpha=0.2))
ax.add_patch(descartes.PolygonPatch(right, fc='r', ec='k', alpha=0.2))
ax.add_patch(descartes.PolygonPatch(middle, fc='g', ec='k', alpha=0.2))
# control display
ax.set_xlim(-2, 2); ax.set_ylim(-2, 2)
ax.set_aspect('equal')
plt.show()
Related
Suppose we have this:
import seaborn as sns
import pandas as pd
import numpy as np
samples = 2**13
data = pd.DataFrame({'Values': list(np.random.normal(size=samples)) + list(np.random.uniform(size=samples)),
'Kind': ['Normal'] * samples + ['Uniform'] * samples})
sns.displot(data, hue='Kind', x='Values', fill=True)
I want my Normal's histogram (or KDE) emphasized. I'd like it in red and non transparent in the background. Uniform should have alpha = .5.
How do I specify these style parameters in a "per hue" manner?
It's possible to do it with two separate histplots on the same Axes, as #Redox suggested. We can basically recreate the same plot, but with fine-grade control over colours and alpha. However I had to explicitly pass the number of bins in to get the same plot as yours. I also needed to define the colour for Uniform otherwise a ghost element would be added to the legend! I used C1, meaning the first default colour.
_, ax = plt.subplots()
sns.histplot(data=data[data.Kind=='Normal'], x="Values", ax=ax, label='Normal', color='tab:red',bins=130,alpha=1)
sns.histplot(data=data[data.Kind=='Uniform'], x="Values", ax=ax, label='Uniform', color='C1',bins=17, alpha=.5)
ax.set_xlabel('')
ax.legend()
Note that if you just want to set the colour without alpha you can already do this on a displot via the palette argument - pass in a dictionary of your unique hue values to colour names. However, the alpha that you pass in must be a scalar. I tried to use this clever answer to set colours as RGBA colours which include alpha, which seems to work with other figure level plots in Seaborn. However, displot overrides this and sets the alpha separately!
I want to create a bar-plot in python. I want this plot to be beautiful though and I don't like the looks of python's axes.bar() function. Therefore, I have decided to use plt.vlines(). The challenge here is that my x-data is a list that contains strings and not numerical data. When I plot my graph, the spacing between the two columns (in my example column 2 = 0) is pretty big:
Furthermore, I want a grid. However, I would like to have minor grid lines as well. I know how to get all of this if my data was numerical. But since my x-data contains strings, I don't know how to set x_max. Any suggestions?
Internally, the positions of the labels are numbered 0,1,... So setting the x-limits a bit before 0 and after the last, shows them more centered.
Usually, bars are drawn with their 'feet' on the ground, which can be set via plt.ylim(0, ...). Minor ticks can be positioned for example at multiples of 0.2. Setting the length of the ticks to zero lets the position count for the grid, but suppresses the tick mark.
from matplotlib import pyplot as plt
from matplotlib.ticker import MultipleLocator
import numpy as np
labels = ['Test 1', 'Test 2']
values = [1, 0.7]
fig, ax = plt.subplots()
plt.vlines(labels, 0, values, colors='dodgerblue', alpha=.4, lw=7)
plt.xlim(-0.5, len(labels) - 0.5) # add some padding left and right of the bars
plt.ylim(0, 1.1) # bars usually have their 0 at the bottom
ax.xaxis.set_minor_locator(MultipleLocator(.2))
plt.tick_params(axis='x', which='both', length=0) # ticks not shown, but position serves for gridlines
plt.grid(axis='both', which='both', ls=':') # optionally set the linestyle of the grid
plt.show()
Say there are two datasets: a big "background" set, and much smaller "foreground" set. The foreground set comes from the background, but might be much smaller.
I am interested in showing the entire background distribution in an ordered sns.barplot, and have the foreground set a brighter contrasting color to draw attention to these samples.
The best solution I could find is to display one graph on top of the other, but what happens is the graph shrinks down to the smaller domain. Here's what I mean:
import matplotlib.pyplot as plt
import seaborn
# Load the example car crash dataset
crashes = sns.load_dataset("car_crashes").sort_values("total", ascending=False)
# states of interest
txcahi = crashes[crashes['abbrev'].isin(['TX','CA','HI'])]
# Plot the total crashes
f, ax = plt.subplots(figsize=(10, 5))
plt.xticks(rotation=90, fontsize=10)
sns.barplot(y="total", x="abbrev", data=crashes, label="Total", color="lightgray")
# overlay special states onto gray plot as red bars
sns.barplot(y="total", x="abbrev", data=txcahi, label="Total", color="red")
sns.despine(left=True, bottom=True)
This data produces:
But it should look like this (ignore stylistic differences):
Why doesn't this approach work, and what would be a better way to accomplish this?
A seaborn barplot just plots the its n data along the values of 0 to n-1. If instead you'd use a matplotlib bar plot, which is unit aware (from matplotlib 2.2 onwards), it'll work as expected.
import matplotlib.pyplot as plt
import seaborn as sns
# Load the example car crash dataset
crashes = sns.load_dataset("car_crashes").sort_values("total", ascending=False)
# states of interest
txcahi = crashes[crashes['abbrev'].isin(['TX','CA','HI'])]
# Plot the total crashes
f, ax = plt.subplots(figsize=(10, 5))
plt.xticks(rotation=90, fontsize=10)
plt.bar(height="total", x="abbrev", data=crashes, label="Total", color="lightgray")
plt.bar(height="total", x="abbrev", data=txcahi, label="Total", color="red")
sns.despine(left=True, bottom=True)
I've drawn a plot that looks something like the following:
It was created using the following code:
import numpy as np
import pandas as pd
import matplotlib as mpl
import matplotlib.pyplot as plt
# 1. Plot a figure consisting of 3 separate axes
# ==============================================
plotNames = ['Plot1','Plot2','Plot3']
figure, axisList = plt.subplots(len(plotNames), sharex=True, sharey=True)
tempDF = pd.DataFrame()
tempDF['date'] = pd.date_range('2015-01-01','2015-12-31',freq='D')
tempDF['value'] = np.random.randn(tempDF['date'].size)
tempDF['value2'] = np.random.randn(tempDF['date'].size)
for i in range(len(plotNames)):
axisList[i].plot_date(tempDF['date'],tempDF['value'],'b-',xdate=True)
# 2. Create a new single axis in the figure. This new axis sits over
# the top of the axes drawn previously. Make all the components of
# the new single axis invisibe except for the x and y labels.
big_ax = figure.add_subplot(111)
big_ax.set_axis_bgcolor('none')
big_ax.set_xlabel('Date',fontweight='bold')
big_ax.set_ylabel('Random normal',fontweight='bold')
big_ax.tick_params(labelcolor='none', top='off', bottom='off', left='off', right='off')
big_ax.spines['right'].set_visible(False)
big_ax.spines['top'].set_visible(False)
big_ax.spines['left'].set_visible(False)
big_ax.spines['bottom'].set_visible(False)
# 3. Plot a separate figure
# =========================
figure2,ax2 = plt.subplots()
ax2.plot_date(tempDF['date'],tempDF['value2'],'-',xdate=True,color='green')
ax2.set_xlabel('Date',fontweight='bold')
ax2.set_ylabel('Random normal',fontweight='bold')
# Save plot
# =========
plt.savefig('tempPlot.png',dpi=300)
Basically, the rationale for plotting the whole picture is as follows:
Create the first figure and plot 3 separate axes using a loop
Plot a single axis in the same figure to sit on top of the graphs
drawn previously. Label the x and y axes. Make all other aspects of
this axis invisible.
Create a second figure and plot data on a single axis.
The plot displays just as I want when using jupyter-notebook but when the plot is saved, the file contains only the second figure.
I was under the impression that plots could have multiple figures and that figures could have multiple axes. However, I suspect I have a fundamental misunderstanding of the differences between plots, subplots, figures and axes. Can someone please explain what I'm doing wrong and explain how to get the whole image to save to a single file.
Matplotlib does not have "plots". In that sense,
plots are figures
subplots are axes
During runtime of a script you can have as many figures as you wish. Calling plt.save() will save the currently active figure, i.e. the figure you would get by calling plt.gcf().
You can save any other figure either by providing a figure number num:
plt.figure(num)
plt.savefig("output.png")
or by having a refence to the figure object fig1
fig1.savefig("output.png")
In order to save several figures into one file, one could go the way detailed here: Python saving multiple figures into one PDF file.
Another option would be not to create several figures, but a single one, using subplots,
fig = plt.figure()
ax = plt.add_subplot(611)
ax2 = plt.add_subplot(612)
ax3 = plt.add_subplot(613)
ax4 = plt.add_subplot(212)
and then plot the respective graphs to those axes using
ax.plot(x,y)
or in the case of a pandas dataframe df
df.plot(x="column1", y="column2", ax=ax)
This second option can of course be generalized to arbitrary axes positions using subplots on grids. This is detailed in the matplotlib user's guide Customizing Location of Subplot Using GridSpec
Furthermore, it is possible to position an axes (a subplot so to speak) at any position in the figure using fig.add_axes([left, bottom, width, height]) (where left, bottom, width, height are in figure coordinates, ranging from 0 to 1).
I'm trying to create a figure like this one from the seaborn documentation but with the edgecolor of the stripplot determined by the hue. This is my attempt:
import seaborn as sns
df = sns.load_dataset("tips")
ax = sns.stripplot(x="sex", y="tip", hue="day", data=df, jitter=True,
edgecolor=sns.color_palette("hls", 4),
facecolors="none", split=False, alpha=0.7)
But the color palettes for male and female appear to be different. How do I use the same color palette for both categories?
I'm using seaborn 0.6.dev
The edgecolor parameter is just passed straight through to plt.scatter. Currently you're giving it a list of 4 colors. I'm not exactly sure what I would expect it to do in that case (and I am not exactly sure why you end up with what you're seeing here), but I would not have expected it to "work".
The ideal way to have this work would be to have a "hollow circle" marker glyph that colors the edges based on the color (or facecolor) attribute rather than the edges. While it would be nice to have this as an option in core matplotlib, there are some inconsistencies that might make that unworkable. However, it's possible to hack together a custom glyph that will do the trick:
import numpy as np
import matplotlib as mpl
import seaborn as sns
sns.set_style("whitegrid")
df = sns.load_dataset("tips")
pnts = np.linspace(0, np.pi * 2, 24)
circ = np.c_[np.sin(pts) / 2, -np.cos(pts) / 2]
vert = np.r_[circ, circ[::-1] * .7]
open_circle = mpl.path.Path(vert)
sns.stripplot(x="sex", y="tip", hue="day", data=df,
jitter=True, split=False,
palette="hls", marker=open_circle, linewidth=0)
FWIW I should also mention that it's important to be careful when using this approach because the colors become much harder to distinguish. The hls palette exacerbates the problem as the lime green and cyan middle colors end up quite similar. I can imagine situations where this would work nicely, though, for instance a hue variable with two levels represented by gray and a bright color, where you want to emphasize the latter.