I'm using Spyder v5 within Anaconda and having issues with any styles being applied on charts. Basic chart below (deliberately commented out the sns import to see if it makes any difference -it doesn't)
Basic code below - gives a normal plot as expected
import matplotlib.pyplot as plt
#import seaborn as sns
x_values = [0,1,2,3,4,5,6,7,8,9,10]
x_squared = [x ** 2 for x in x_values]
plt.plot(x_values,x_squared, label = "X-Squared")
plt.plot(x_values,x_cubed, label = "X-Cubed")
plt.show()
If i then try
plt.style.use("seaborn-deep")
plt.plot(x_values,x_squared, label = "X-Squared")
plt.plot(x_values,x_cubed, label = "X-Cubed")
plt.show()
Nothing changes? The style is avaialble (from plt.styles.available) and there is no error when applying the style so something is amiss here.
If i import seaborn then nothing changes
If i try any or all of the following from other solutions in various places in my plt code then nothing much changes (all of the below lines of code work with no errors
%matplotlib inline
plt.rcParams.update(plt.rcParamsDefault)
sns.set()
One time somehow this did seem to force it to change to a seaborn-paper style but that's then the only one i can use and i can't use any other seaborn or other styles at all?? Not sure what forced it to change, think it was sns.set() so I suspect the one it then shows is the default seaborn style, but then why won't it let me use any other styles?
As i'm using SPyder in anaconda the packages are installed to do this plot
I do have Python installed on my base desktop as well... but documentation insists this won't cause a problem and i've not had this problem on any other code ever of having another installation
any tips?
Related
I finished analyzing my data and want to show that they are statistically significant using the t-test_ind. However, I haven't found anything functional to show this other than what was referenced in (How does one insert statistical annotations (stars or p-values) into matplotlib / seaborn plots?):
import seaborn as sns
import pandas as pd
import matplotlib.pyplot as plt
from statannot import add_stat_annotation
ax = sns.barplot(x=x, y=y, order=order)
add_stat_annotation(ax, data=df, x=x, y=y,
boxPairList=[(order[0], order[1]), (order[0], order[2])],
test='t-test_ind',
textFormat='star',
loc='outside')
Using this approach however, whenever I try to save the plot using plt.savefig() the added significancies using the add_stat_annotation are discared (matplotlib does not seem to recognize the added annotations). Using the loc='inside' option messes up my plot so it isn't really an option.
I am therefore asking if there is some simpler way to add the sigificancies directly in matplotlib / seaborn or if you can plt.savefig() with enough border / padding to include everything.
It was mainly a xlabel cut off problem. So in future applications I would use the add_stat_annotation from webermarcolivier/statannot. To save your files use one of the following possibilities:
import matplotlib.pyplot as plt
plt.tight_layout() # Option 1
plt.autoscale() # Option 2
plt.savefig('filename.png', bbox_inches = "tight") # Option 3
Hope this will help someone for future use.
I am trying to learn python mainly for plotting. Here is my sample code:
import numpy as np
import matplotlib.pyplot as plt
a=[[1,2,3,4],[2,3,4,5],[3,4,5,6]]
x=np.arange(len(a[0]))
width=0.2
fig, ax = plt.subplots(figsize=(8,6))
patterns=['/','\\','*']
for bar in a:
ax.bar(x,bar,width,edgecolor='black',color='lightgray', hatch=patterns.pop(0))
x=x+width
plt.show()
Now the problem is that, I need black edge colour for all bars as well as given hatch patter. However, the formatting is applied to first set of bars only. Here is my output. (I am using python3).
What's missing here or what's wrong? I have looked around but did not find any fix.
Update:
I have tried different options :python2, python3 and pdf/png. Here are results
python2 png --fine
python3 png -- shown above
python2 pdf -- see
python3 pdf -- see
I have also tried 'backend' as matplotlib.use('Agg'). I have update my matplotlib version (2.1.0).
There is a current issue in matplotlib 2.1 that only the first bar's edgecolor is applied. The same for the hatch, see this issue. Also see this question.
It may be that you are using matplotlib 2.1 for python3 but not for python2, hence in python2 it works for you. If I run your code in python 2 with matplotlib 2.1 I get the same undesired behaviour.
The issue will be fixed, once matplotlib 2.1.1 is released.
In the meantime, a workaround is to set the edgecolor and hatch on the individual bars:
import numpy as np
import matplotlib.pyplot as plt
a=[[1,2,3,4],[2,3,4,5],[3,4,5,6]]
x=np.arange(len(a[0]))
width=0.2
fig, ax = plt.subplots(figsize=(8,6))
patterns=['/','\\','*']
for y in a:
bars = ax.bar(x,y,width,color='lightgray')
hatch= patterns.pop(0)
for bar in bars:
bar.set_edgecolor("black")
bar.set_hatch(hatch)
x=x+width
plt.show()
It looks something's wrong with edgecolor tuple's alpha value. Set it to 1 will solve the problem.
My plots in matplotlib don't show more than one letter/digit.
I do need to use matplotlib from the anaconda folder which is why I'm including sys.path at the beginning. (When I don't include the first 2 lines everything works fine since I have a different matplotlib on my computer. Both are up to date.) Any ideas what I can do to make it work? I'm using Ubuntu and Sublime Text.
Here's my simple code:
import sys
sys.path.insert(0, "/home/mariusz/anaconda/lib/python2.7/site-packages")
import matplotlib.pyplot as plt
grade_fast = [1,7,3,5,6,7,10,11,12]
bumpy_fast = [0,2,3,5,6,7,10,11,12]
grade_slow = [0,2,3,5,6,7,10,11,12]
bumpy_slow = [0,2,3,5,6,7,10,11,12]
plt.scatter(bumpy_fast, grade_fast, label="f")
plt.scatter(grade_slow, bumpy_slow, label="slow")
plt.legend()
plt.xlabel("bumpiness")
plt.ylabel("g")
plt.show()
I re-installed matplotlib, numpy, scikit-learn and scipy and it worked :)
I don't know why though.
In the past I was able to do simple animations with matplotlib with a for loop, but this hasn't worked for some time now.
The standard answer is that you have to turn interactive mode on and/or force a redraw with matplotlib.pyplot.draw(). Here is my minimal working example:
import numpy as np
import matplotlib
matplotlib.use('Qt4Agg')
import matplotlib.pyplot as mplot
mplot.ion()
fig = mplot.figure(1)
ax = fig.add_subplot(111)
for ii in np.arange(0,10):
x = 200*np.random.rand(30)
ax.plot(x)
mplot.draw()
filename = ("img_%d.png" % ii)
mplot.savefig(filename)
When I run this in Interactive Python Editor, I get one figure at the very end with all the plots in it (this also happens with mplot.show())
When I run this in IPython 3.1 (with Python 3.3.5) from the command line, I get nothing at all.
The mplot.savefig(filename) line does seem to work, as the images are generated.
(It's possible this is a bug in the Qt4 backend.)
Try deleting the line matplotlib.use('Qt4Agg'). Works for me. Also works with matplotlib.use('TkAgg'). So it is a backend problem. There is another way to do animations.
I'm brand new to Python, I just switched from Matlab. The distro is Anaconda 2.1.0 and I'm using the Spyder IDE that came with it.
I'm trying to make a scatter plot with equal ratios on the x and y axes, so that this code prints a square figure with the vertices of a regular hexagon plotted inside.
import numpy
import cmath
import matplotlib
coeff = [1,0,0,0,0,0,-1]
x = numpy.roots(coeff)
zeroplot = plot(real(x),imag(x), 'ro')
plt.gca(aspect='equal')
plt.show()
But plt.gca(aspect='equal') returns a blank figure with axes [0,1,0,1], and plt.show() returns nothing.
I think the main problem is that plt.gca(aspect='equal') doesn't just grab the current axis and set its aspect ratio. From the documentation, (help(plt.gca)) it appears to create a new axis if the current one doesn't have the correct aspect ratio, so the immediate fix for this should be to replace plt.gca(aspect='equal') with:
ax = plt.gca()
ax.set_aspect('equal')
I should also mention that I had a little bit of trouble getting your code running because you're using pylab to automatically load numpy and matplotlib functions: I had to change my version to:
import numpy
import cmath
from matplotlib import pyplot as plt
coeff = [1,0,0,0,0,0,-1]
x = numpy.roots(coeff)
zeroplot = plt.plot(numpy.real(x), numpy.imag(x), 'ro')
ax = plt.gca()
ax.set_aspect('equal')
plt.show()
People who are already comfortable with Python don't generally use Pylab, from my experience. In future you might find it hard to get help on things if people don't realise that you're using Pylab or aren't familiar with how it works. I'd recommend disabling it and trying to get used to accessing the functions you need through their respective modules (e.g. using numpy.real instead of just real)