I have 3 1-D ndarrays: x, y, z
and the following code:
import numpy as np
import matplotlib.pyplot as plt
import scipy.interpolate as spinterp
## define data
npoints = 50
xreg = np.linspace(x.min(),x.max(),npoints)
yreg = np.linspace(y.min(),y.max(),npoints)
X,Y = np.meshgrid(xreg,yreg)
Z = spinterp.griddata(np.vstack((x,y)).T,z,(X,Y),
method='linear').reshape(X.shape)
## plot
plt.close()
ax = plt.axes()
col = ax.pcolormesh(X,Y,Z.T)
plt.draw()
My plot comes out blank, and I suspect it is because the method='linear' interpolation comes out with nans. I've tried converting to a masked array, but to no avail - plot is still blank. Can you tell me what I am doing wrong? Thanks.
Got it. This seems round-about, but this was the solution:
import numpy.ma as ma
Zm = ma.masked_where(np.isnan(Z),Z)
plt.pcolormesh(X,Y,Zm.T)
If the Z matrix contains nan's, it has to be a masked array for pcolormesh, which has to be created with ma.masked_where, or, alternatively,
Zm = ma.array(Z,mask=np.isnan(Z))
A slight improvement on the chosen answer
import numpy.ma as ma
Zm = ma.masked_invalid(Z)
plt.pcolormesh(X, Y, Zm.T)
masked_invalid masks all NaN values, thereby saving the need to specify
mask = np.isnan(Z)
Note that the explicit masking is no longer necessary in matplotlib master as arrays are now masked automatically internally. Will be incorporated into matplotlib >2.1. See my merged pull request https://github.com/matplotlib/matplotlib/pull/5451
So now it's as simple as
plt.pcolormesh(X,Y,Z.T)
Related
I have two data sets index_list and frequency_list which I plot in a loglog plot by plt.loglog(index_list, freq_list). Now I'm trying to fit a power law a*x^(-b) with linear regression. I expect the curve to follow the initial curve closely but the following code seems to output a similar curve but mirrored on the y-axis.
I suspect I am using curve_fit badly.
why is this curve mirrored on the x-axis and how I can get it to properly fit my inital curve?
Using this data
import numpy as np
import matplotlib.pyplot as plt
from scipy.optimize import curve_fit
f = open ("input.txt", "r")
index_list = []
freq_list = []
index = 0
for line in f:
split_line = line.split()
freq_list.append(int(split_line[1]))
index_list.append(index)
index += 1
plt.loglog(index_list, freq_list)
def power_law(x, a, b):
return a * np.power(x, -b)
popt, pcov = curve_fit(power_law, index_list, freq_list)
plt.plot(index_list, power_law(freq_list, *popt))
plt.show()
The code below made the following changes:
For the scipy functions to work, it is best that both index_list and freq_list are numpy arrays, not Python lists. Also, for the power not to overflow too rapidly, these arrays should be of float type (not of int).
As 0 to a negative power causes a divide-by-zero problem, it makes sense to start the index_list with 1.
Due to the powers, also for floats an overflow can be generated. Therefore, it makes sense to add bounds to curve_fit. Especially b should be limited not to cross about 50 (the highest value is about power(100000, b) giving an overflow when be.g. is100). Also setting initial values helps to direct the fitting process (p0=...).
Drawing a plot with index_list as x and power_law(freq_list, ...) as y would generate a very weird curve. It is necessary that the same x is used for the plot and for the function.
Note that calling plt.loglog() changes both axes of the plot to logarithmic. All subsequent plots on the same axes will continue to use the logarithmic scale.
import matplotlib.pyplot as plt
from scipy.optimize import curve_fit
import pandas as pd
import numpy as np
def power_law(x, a, b):
return a * np.power(x, -b)
df = pd.read_csv("https://norvig.com/google-books-common-words.txt", delim_whitespace=True, header=None)
index_list = df.index.to_numpy(dtype=float) + 1
freq_list = df[1].to_numpy(dtype=float)
plt.loglog(index_list, freq_list, label='given data')
popt, pcov = curve_fit(power_law, index_list, freq_list, p0=[1, 1], bounds=[[1e-3, 1e-3], [1e20, 50]])
plt.plot(index_list, power_law(index_list, *popt), label='power law')
plt.legend()
plt.show()
As shown in the figure,
How can I plot a line that have different colors based on a specific value of x ?
The simplest solution here may be to slice your data at the corresponding index of x_lim found by np.where :
from matplotlib import pyplot as plt
import numpy as np
x = np.linspace(0,2*np.pi,100)
y = np.cos(x)*np.exp(-x/2)
# specify your x limitation
x_lim = np.pi
# find the first corresponding idx where the condition x>=x_lim hold
x_lim_idx = np.where(x>=x_lim)[0][0]
# plot sliced data
plt.plot(x[:x_lim_idx],y[:x_lim_idx],'r')
plt.plot(x[x_lim_idx:],y[x_lim_idx:],'b')
which gives for x_lim = np.pi :
And if the remaining gap between the lines bothers you, for small x discretization for instance, you can still close it by making the two slices overlap.
I'd like to make a streamplot with lines that don't stop when they get too close together. I'd rather each streamline be calculated in both directions until it hits the edge of the window. The result is there'd be some areas where they'd all jumble up. But that's what I want.
I there anyway to do this in matplotlib? If not, is there another tool I can use for this that could interface with python/numpy?
import numpy as np
import matplotlib.pyplot as plt
Y,X = np.mgrid[-10:10:.01, -10:10:.01]
U, V = Y**2, X**2
plt.streamplot(X,Y, U,V, density=1)
plt.show(False)
Ok, I've figured out I can get mostly what I want by turning up the density a lot and using custom start points. I'm still interested if there is a better or alternate way to do this.
Here's my solution. Doesn't it look so much better?
import numpy as np
import matplotlib.pyplot as plt
Y,X = np.mgrid[-10:10:.01, -10:10:.01]
y,x = Y[:,0], X[0,:]
U, V = Y**2, X**2
stream_points = np.array(zip(np.arange(-9,9,.5), -np.arange(-9,9,.5)))
plt.streamplot(x,y, U,V, start_points=stream_points, density=35)
plt.show(False)
Edit: By the way, there seems to be some bug in streamplot such that start_points keyword only works if you use 1d arrays for the grid data. See Python Matplotlib Streamplot providing start points
As of Matplotlib version 3.6.0, an optional parameter broken_streamlines has been added for disabling streamline breaks.
Adding it to your snippet produces the following result:
import numpy as np
import matplotlib.pyplot as plt
Y,X = np.mgrid[-10:10:.01, -10:10:.01]
U, V = Y**2, X**2
plt.streamplot(X,Y, U,V, density=1, broken_streamlines=False)
plt.show(False)
Note
This parameter just extends the streamlines which were originally drawn (as in the question). This means that the streamlines in the modified plot above are much more uneven than the result obtained in the other answer, with custom start_points. The density of streamlines on any stream plot does not represent the magnitude of U or V at that point, only their direction. See the documentation for the density parameter of matplotlib.pyplot.streamplot for more details on how streamline start points are chosen by default, when they aren't specified by the optional start_points parameter.
For accurate streamline density, consider using matplotlib.pyplot.contour, but be aware that contour does not show arrows.
Choosing start points automatically
It may not always be easy to choose a set of good starting points automatically. However, if you know the streamfunction corresponding to the flow you wish to plot you can use matplotlib.pyplot.contour to produce a contour plot (which can be hidden from the output), and then extract a suitable starting point from each of the plotted contours.
In the following example, psi_expression is the streamfunction corresponding to the flow. When modifying this example for your own needs, make sure to update both the line defining psi_expression, as well as the one defining U and V. Ensure these both correspond to the same flow.
The density of the streamlines can be altered by changing contour_levels. Here, the contours are uniformly distributed.
import numpy as np
import matplotlib.pyplot as plt
import sympy as sy
x, y = sy.symbols("x y")
psi_expression = x**3 - y**3
psi_function = sy.lambdify((x, y), psi_expression)
Y, X = np.mgrid[-10:10:0.01, -10:10:0.01]
psi_evaluated = psi_function(X, Y)
U, V = Y**2, X**2
contour_levels = np.linspace(np.amin(psi_evaluated), np.amax(psi_evaluated), 30)
# Draw a temporary contour plot.
temp_figure = plt.figure()
contour_plot = plt.contour(X, Y, psi_evaluated, contour_levels)
plt.close(temp_figure)
points_list = []
# Iterate over each contour.
for collection in contour_plot.collections:
# Iterate over each segment in this contour.
for path in collection.get_paths():
middle_point = path.vertices[len(path.vertices) // 2]
points_list.append(middle_point)
# Reshape python list into numpy array of coords.
stream_points = np.reshape(np.array(points_list), (-1, 2))
plt.streamplot(X, Y, U, V, density=1, start_points=stream_points, broken_streamlines=False)
plt.show(False)
I have an existing plot that was created with pandas like this:
df['myvar'].plot(kind='bar')
The y axis is format as float and I want to change the y axis to percentages. All of the solutions I found use ax.xyz syntax and I can only place code below the line above that creates the plot (I cannot add ax=ax to the line above.)
How can I format the y axis as percentages without changing the line above?
Here is the solution I found but requires that I redefine the plot:
import matplotlib.pyplot as plt
import numpy as np
import matplotlib.ticker as mtick
data = [8,12,15,17,18,18.5]
perc = np.linspace(0,100,len(data))
fig = plt.figure(1, (7,4))
ax = fig.add_subplot(1,1,1)
ax.plot(perc, data)
fmt = '%.0f%%' # Format you want the ticks, e.g. '40%'
xticks = mtick.FormatStrFormatter(fmt)
ax.xaxis.set_major_formatter(xticks)
plt.show()
Link to the above solution: Pyplot: using percentage on x axis
This is a few months late, but I have created PR#6251 with matplotlib to add a new PercentFormatter class. With this class you just need one line to reformat your axis (two if you count the import of matplotlib.ticker):
import ...
import matplotlib.ticker as mtick
ax = df['myvar'].plot(kind='bar')
ax.yaxis.set_major_formatter(mtick.PercentFormatter())
PercentFormatter() accepts three arguments, xmax, decimals, symbol. xmax allows you to set the value that corresponds to 100% on the axis. This is nice if you have data from 0.0 to 1.0 and you want to display it from 0% to 100%. Just do PercentFormatter(1.0).
The other two parameters allow you to set the number of digits after the decimal point and the symbol. They default to None and '%', respectively. decimals=None will automatically set the number of decimal points based on how much of the axes you are showing.
Update
PercentFormatter was introduced into Matplotlib proper in version 2.1.0.
pandas dataframe plot will return the ax for you, And then you can start to manipulate the axes whatever you want.
import pandas as pd
import numpy as np
df = pd.DataFrame(np.random.randn(100,5))
# you get ax from here
ax = df.plot()
type(ax) # matplotlib.axes._subplots.AxesSubplot
# manipulate
vals = ax.get_yticks()
ax.set_yticklabels(['{:,.2%}'.format(x) for x in vals])
Jianxun's solution did the job for me but broke the y value indicator at the bottom left of the window.
I ended up using FuncFormatterinstead (and also stripped the uneccessary trailing zeroes as suggested here):
import pandas as pd
import numpy as np
from matplotlib.ticker import FuncFormatter
df = pd.DataFrame(np.random.randn(100,5))
ax = df.plot()
ax.yaxis.set_major_formatter(FuncFormatter(lambda y, _: '{:.0%}'.format(y)))
Generally speaking I'd recommend using FuncFormatter for label formatting: it's reliable, and versatile.
For those who are looking for the quick one-liner:
plt.gca().set_yticklabels([f'{x:.0%}' for x in plt.gca().get_yticks()])
this assumes
import: from matplotlib import pyplot as plt
Python >=3.6 for f-String formatting. For older versions, replace f'{x:.0%}' with '{:.0%}'.format(x)
I'm late to the game but I just realize this: ax can be replaced with plt.gca() for those who are not using axes and just subplots.
Echoing #Mad Physicist answer, using the package PercentFormatter it would be:
import matplotlib.ticker as mtick
plt.gca().yaxis.set_major_formatter(mtick.PercentFormatter(1))
#if you already have ticks in the 0 to 1 range. Otherwise see their answer
I propose an alternative method using seaborn
Working code:
import pandas as pd
import seaborn as sns
data=np.random.rand(10,2)*100
df = pd.DataFrame(data, columns=['A', 'B'])
ax= sns.lineplot(data=df, markers= True)
ax.set(xlabel='xlabel', ylabel='ylabel', title='title')
#changing ylables ticks
y_value=['{:,.2f}'.format(x) + '%' for x in ax.get_yticks()]
ax.set_yticklabels(y_value)
You can do this in one line without importing anything:
plt.gca().yaxis.set_major_formatter(plt.FuncFormatter('{}%'.format))
If you want integer percentages, you can do:
plt.gca().yaxis.set_major_formatter(plt.FuncFormatter('{:.0f}%'.format))
You can use either ax.yaxis or plt.gca().yaxis. FuncFormatter is still part of matplotlib.ticker, but you can also do plt.FuncFormatter as a shortcut.
Based on the answer of #erwanp, you can use the formatted string literals of Python 3,
x = '2'
percentage = f'{x}%' # 2%
inside the FuncFormatter() and combined with a lambda expression.
All wrapped:
ax.yaxis.set_major_formatter(FuncFormatter(lambda y, _: f'{y}%'))
Another one line solution if the yticks are between 0 and 1:
plt.yticks(plt.yticks()[0], ['{:,.0%}'.format(x) for x in plt.yticks()[0]])
add a line of code
ax.yaxis.set_major_formatter(ticker.PercentFormatter())
When plotting using matplotlib, I ran into an interesting issue where the y axis is scaled by a very inconvenient quantity. Here's a MWE that demonstrates the problem:
import numpy as np
import matplotlib.pyplot as plt
l = np.linspace(0.5,2,2**10)
a = (0.696*l**2)/(l**2 - 9896.2e-9**2)
plt.plot(l,a)
plt.show()
When I run this, I get a figure that looks like this picture
The y-axis clearly is scaled by a silly quantity even though the y data are all between 1 and 2.
This is similar to the question:
Axis numerical offset in matplotlib
I'm not satisfied with the answer to this question in that it makes no sense to my why I need to go the the convoluted process of changing axis settings when the data are between 1 and 2 (EDIT: between 0 and 1). Why does this happen? Why does matplotlib use such a bizarre scaling?
The data in the plot are all between 0.696000000017 and 0.696000000273. For such cases it makes sense to use some kind of offset.
If you don't want that, you can use you own formatter:
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.ticker
l = np.linspace(0.5,2,2**10)
a = (0.696*l**2)/(l**2 - 9896.2e-9**2)
plt.plot(l,a)
fmt = matplotlib.ticker.StrMethodFormatter("{x:.12f}")
plt.gca().yaxis.set_major_formatter(fmt)
plt.show()