notebook.community



In [1]:

    
import matplotlib as mpl
mpl.use('TkAgg')
import matplotlib.pyplot as plt
%matplotlib inline

import pandas as pd
import seaborn as sns



In [2]:

    
import sys
print(sys.version)









    



3.5.2 |Continuum Analytics, Inc.| (default, Jul  2 2016, 17:52:12) 
[GCC 4.2.1 Compatible Apple LLVM 4.2 (clang-425.0.28)]



In [3]:

    
!python --version









    



Python 3.5.2 :: Continuum Analytics, Inc.

Prepare a minimal data set.



In [4]:

    
dataframe = pd.DataFrame({'age':[1.,2,3,4,5,6,7,8,9],
                         'height':[4, 4.5, 5, 6, 7, 8, 9, 9.5, 10],
                         'gender':['M','F', 'F','M','M','F', 'F','M', 'F'],
                         #'hair color':['brown','black', 'brown', 'blonde', 'brown', 'red', 
                         #              'brown', 'brown', 'black' ],
                         'hair length':[1,6,2,3,1,5,6,5,3] })



In [5]:

    
dataframe









    Out[5]:






  
    
      
      age
      gender
      hair length
      height
    
  
  
    
      0
      1.0
      M
      1
      4.0
    
    
      1
      2.0
      F
      6
      4.5
    
    
      2
      3.0
      F
      2
      5.0
    
    
      3
      4.0
      M
      3
      6.0
    
    
      4
      5.0
      M
      1
      7.0
    
    
      5
      6.0
      F
      5
      8.0
    
    
      6
      7.0
      F
      6
      9.0
    
    
      7
      8.0
      M
      5
      9.5
    
    
      8
      9.0
      F
      3
      10.0

You can pass seaborn color palettes of RGB values.

For example:



In [6]:

    
sns.color_palette(palette="Blues_d", n_colors=9, desat=None)









    Out[6]:





[(0.20307574244885662, 0.26459054565118034, 0.30867615307078644),
 (0.20627451459566754, 0.33176471312840783, 0.42169935226440436),
 (0.20935025704452415, 0.3963552587795881, 0.53037550533519073),
 (0.21254902919133503, 0.46352942625681565, 0.64339870452880865),
 (0.21747534201036089, 0.53052930169635348, 0.75482251877878226),
 (0.30692811806996667, 0.58640524546305339, 0.78352942466735842),
 (0.39995900517195659, 0.64451622698042133, 0.81338460679147762),
 (0.48941178123156226, 0.70039217074712123, 0.84209151268005367),
 (0.58244266833355229, 0.75850315226448906, 0.87194669480417286)]

The following plot uses a seaborn-derived color pallette. See below for info about how to make, modify, and view them.



In [7]:

    
def plot_data():
    color_palette = sns.color_palette(palette="Blues_d", n_colors=9, desat=None)
    g = sns.FacetGrid(dataframe, col="gender", hue='height', palette=color_palette)
    g = (g.map(plt.scatter, "age", 'hair length', edgecolor="w").add_legend())

plot_data()



In [8]:

    
# change seaborn style
sns.set(style="ticks")

# re-make the plot with this setting. 
plot_data()

Add some settings to matplotlib to get bigger fonts.

Note: not all of their suggestions are appropriate for my distribution/needs.



In [9]:

    
# settings from I heart pandas: 
# https://github.com/fredhutchio/i-heart-pandas/blob/master/3-pandas.ipynb
mpl.rcParams.update({
    'font.size': 16, 'axes.titlesize': 17, 'axes.labelsize': 15,
    'xtick.labelsize': 10, 'ytick.labelsize': 13,
    #'font.family': 'Lato',  # I don't have this font family
        'font.weight': 600,
    'axes.labelweight': 600, 'axes.titleweight': 600,
    #'figure.autolayout': True  # screws up layout for seaborn facetgrid objects. 
    })



In [10]:

    
# re-make the plot with this setting. 
plot_data()

Making custom color maps

You can pass seaborn custom RGB values. Just put the lists of RGB values in a list:



In [11]:

    
# a sample seaborn-created color palette: 
# http://stanford.edu/~mwaskom/software/seaborn-dev/tutorial/color_palettes.html 

sns.diverging_palette(145, 280, s=85, l=25, n=7)









    Out[11]:





[array([ 0.07106475,  0.2657334 ,  0.17177439,  1.        ]),
 array([ 0.36441028,  0.51184576,  0.44068437,  1.        ]),
 array([ 0.66442276,  0.76355158,  0.71570594,  1.        ]),
 array([ 0.95,  0.95,  0.95,  1.  ]),
 array([ 0.7629515 ,  0.6726194 ,  0.84602388,  1.        ]),
 array([ 0.54270509,  0.38706233,  0.68583934,  1.        ]),
 array([ 0.32735306,  0.10785097,  0.52921445,  1.        ])]

Seaborn also comes with a function called palplot, which can display color palettes.



In [12]:

    
sns.palplot(sns.diverging_palette(145, 280, s=85, l=25, n=7))

Since the palettes are just lists of list, you can path them together as you like.



In [13]:

    
sns.palplot(sns.diverging_palette(145, 280, s=85, l=25, n=7)[0:5] +
            sns.diverging_palette(220, 20, n=7)[4:])



In [ ]:

	age	gender	hair length	height
0	1.0	M	1	4.0
1	2.0	F	6	4.5
2	3.0	F	2	5.0
3	4.0	M	3	6.0
4	5.0	M	1	7.0
5	6.0	F	5	8.0
6	7.0	F	6	9.0
7	8.0	M	5	9.5
8	9.0	F	3	10.0