I. Comparison between Seaborn and Matplotlib
Seaborn is a powerful extension of
Matplotlib.
A case in point
Ask for a scatter plot of the length of the calyx and petal, and color to distinguish the species of flower
There are three varieties of flowers:
Define the color of each flower according to the type of flower
Color_map = dict (zip (iris) Name) unique (), [' blue ', 'green', 'red']))
Use Matplotlib to draw the picture
For species, Group in iris. Group by('Name'):
PLT. Scatter (group [' PetalLength], group [' SepalLength],
Color = color_map [species],
Alpha = 0.3, edgecolor = None,
Label = species)
PLT. Legend (frameon = True, the title = 'Name')
PLT. Xlabel (' petalLength ')
PLT. Ylabel (' sepalLength ')
Draw with Seaborn
Seaborn is much easier than matplotlib to draw a scatter plot with just one line of code.
Nss.lmplot ('PetalLength', 'SepalLength', Iris, Hue ='Name', fit_reg=False)
Second, Seaborn realizes histogram and density map
0x1 recalls the Matplotlib method
S1 = Series (np) random) randn (1000))
PLT. Hist (s1)
= 'kde' s1. The plot (kind)
0X2 draws histograms
Seaborn has a powerful approach: distplot, which supports a few parameters:
Bins: blocks of histogram
Hist: True means drawing histogram, and the default is True
Kde: True means to draw a density graph, and the default is True
Rug: Displays distribution, False by default, not displayed
SNS. Distplot (s1, hist = True, kde = True)
You can see the data distribution below
0x3 draws a density map
The density map can be drawn by directly passing in the data:
You can also specify the color with the color parameter:
SNS. Kdeplot (s1, shade = True, color = 'r')
tip
The PLT function can be called directly through NSS.plT
3. Seaborn realizes the bar chart and thermal diagram
0x1 data preparation
Seaborn provided a load_dataset method which can download data online as an experiment. This method was used to generate experiment data:
Load_dataset implementation source at https://github.com/mwaskom/seaborn/blob/master/seaborn/utils.py
PivotTable
Pivot (index='month', Columns ='year', values='passengers')
0x2 maps the heat map
Seaborn provided the heatMap method for plotting a heatmap:
The parameters annot=True, FMT ='d' can make each square display specific value in the thermal diagram:
0X2 draws a histogram
The abscissa of the bar chart is the year, and the ordinate is the sum of all passengers in the month of the year:
Firstly, the sum method is used to calculate the passenger sum of each year:
Where index is the year and VALUES is the sum of the passengers of the year
Seaborn provides the Barplot method with bar-bar plot, just specifying x and Y coordinates in the parameter:
SNS. Barplot (x = s.i ndex, y = s.v alues)
