Tips: teach you to generate 30 cool dynamic interactive charts with one click of pandas

Today, let's talk about how to use one line of code in DataFrame Generate cool dynamic interactive charts in the dataset , Let's first introduce the modules we need to use this time cufflinks
like seaborn Encapsulates the matplotlib equally ,cufflinks Also in the plotly Further packaging and optimization are made above , The methods are unified 、 Simple parameter configuration , about DataFrame It is also convenient and flexible for drawing data sets , And the chart we're going to draw this time includes

• Broken line diagram
• Area map
• Scatter plot
• Histogram
• Histogram
• Box figure
• Heat map
• 3D Scatter plot /3D Bubble chart
• Trend chart
• The pie chart
• K Line graph
• Multiple subgraphs are combined

Module installation

Involving installation , direct pip install that will do

pip install cufflinks
 Copy code 

The import module , And view the relevant configuration

We import the module , Let's see what the current version is

cf.__version__
 Copy code 

output

'0.17.3'
 Copy code 

At present, the version of the module has reached 0.17.3, It's also the latest version , Then, what charts can be drawn in our latest version

cf.help()
 Copy code 

output

Use 'cufflinks.help(figure)' to see the list of available parameters for the given figure.
Use 'DataFrame.iplot(kind=figure)' to plot the respective figure
Figures:
 bar
 box
 bubble
 bubble3d
 candle
 choroplet
 distplot
 .......
 Copy code 

From the output above we can see , The general syntax for drawing a chart is df.iplot(kind= Chart name ) How do we want to view the parameters when drawing a specific chart , For example, a histogram bar What are the parameters , You can do that

cf.help('bar')
 Copy code 

Histogram

Let's first look at the drawing of histogram chart , First, let's create a data set for chart drawing

df2 = pd.DataFrame({'Category':['A','B','C','D'],
                    'Values':[95,56,70,85]})
df2
 Copy code 

output

  Category  Values
0        A      95
1        B      56
2        C      70
3        D      85
 Copy code 

Then let's draw the histogram

df2.iplot(kind='bar',x='Category',y='Values',
          xTitle = "Category",yTitle = "Values",
          title = " Histogram ")
 Copy code 

output

Among them x The parameter is filled with x The corresponding variable name on the axis , and y The parameter is filled in y The corresponding variable name on the axis , We can draw the chart in png Download it in the format of ,

At the same time, we can also zoom in on the chart ,

Let's take a look at the following set of data

df = pd.DataFrame(np.random.randn(100,4),columns='A B C D'.split())
df.head()
 Copy code 

output

          A         B         C         D
0  0.612403 -0.029236 -0.595502  0.027722
1  1.167609  1.528045 -0.498168 -0.221060
2 -1.338883 -0.732692  0.935410  0.338740
3  1.662209  0.269750 -1.026117 -0.858472
4  1.387077 -0.839192 -0.562382 -0.989672
 Copy code 

Let's plot the histogram

df.head(10).iplot('bar')
 Copy code 

output

We can also draw “ Stacked ” Histogram

df.head(10).iplot(kind='bar',barmode='stack')
 Copy code 

output

So again , We can also draw the histogram horizontally

df.head(10).iplot(kind='barh',barmode='stack')
 Copy code 

output

Broken line diagram

Now let's take a look at the drawing of line chart , Let's start with the above df The columns of the dataset are accumulated

df3 = df.cumsum()
 Copy code 

Then let's draw a line chart

df3.iplot()
 Copy code 

output

Of course, you can also filter out a few columns and draw them , The effect is as follows

df3[["A", "B"]].iplot()
 Copy code 

output

We can also draw a straight line to fit its trend ,

df3['A'].iplot(bestfit = True,bestfit_colors=['pink'])
 Copy code 

output

Here we will focus on introducing a iplot() Parameters commonly used in methods

• kind: Chart type , The default is scatter, Scatter type , There are other types to choose from bar( Histogram )、box( Box figure )、heatmap( Heat map ) wait
• theme: Layout theme , Can pass cf.getThemes() To see what are the main
• title: The title of the chart
• xTitle/yTitle: x perhaps y The name of the shaft above the shaft
• colors: The color when drawing the chart
• subplots: Boolean value , When drawing subgraphs, you need , The default is False
• mode:  character string , Drawing mode , There can be lines、markers, There's also lines+markers and lines+text Equal mode
• size: For scatter charts , It is mainly used to adjust the size of scatter points
• shape: When drawing subgraphs, the layout of each graph
• bargap: The distance between columns in the histogram
• barmode : The shape of histogram ,stack( Stacked )、group( Clusters )、overlay( Cover )

Area map

The transition from line graph to area graph is very simple , Only the parameters fill Set to True that will do , The code is as follows

df3.iplot(fill = True)
 Copy code 

output

Scatter plot

For the drawing of scatter diagram , We need to mode Set to marker, The code is as follows

df3.iplot(kind='scatter',x='A',y='B',
          mode='markers',size=10)
 Copy code 

output

We can adjust size Parameter to adjust the size of the scatter , For example, we will size Adjust to 20

df3.iplot(kind='scatter',x='A',y='B',
          mode='markers',size=20)
 Copy code 

output

Or will mode Set to lines+markers, The code is as follows

df3.iplot(kind='scatter',x='A',y='B',
          mode='lines + markers',size=10)
 Copy code 

We can also set the shape of the scatter , For example, the following code

df3.iplot(kind='scatter',x='A',y='B',
          mode='markers',size=20,symbol="x",
          colorscale='paired',)
 Copy code 

output

Of course, we can also set the color of scatter

df.iplot(kind='scatter' ,mode='markers',
         symbol='square',colors=['orange','purple','blue','red'],
         size=20)
 Copy code 

output

Bubble chart

The presentation of bubble chart is similar to that of scatter chart , On the drawing, you will kind Change the parameter to bubble, Suppose we have such a set of data

cf.datagen.bubble(prefix='industry').head()
 Copy code 

output

          x         y  size    text categories
0  0.332274  1.053811     2  LCN.CG  industry1
1 -0.856835  0.422373    87  ZKY.XC  industry1
2 -0.818344 -0.167020    72  ZSJ.DJ  industry1
3 -0.720254  0.458264    11  ONG.SM  industry1
4 -0.004744  0.644006    40  HUW.DN  industry1
 Copy code 

Let's draw a bubble chart

cf.datagen.bubble(prefix='industry').iplot(kind='bubble',x='x',y='y',size='size',
       categories='categories',text='text', xTitle='Returns',
       yTitle='Analyst Score',title='Cufflinks -  Bubble chart ')
 Copy code 

output

The difference between bubble chart and scatter chart is , The size of each point in the scatter diagram is the same , But the bubble chart is not so

3D Scatter plot

Now that we've mentioned the bubble chart , that 3D By the way, the scatter chart , Suppose our data are as follows

cf.datagen.scatter3d(2,150).head()
 Copy code 

output

          x         y         z    text categories
0  0.375359 -0.683845 -0.960599  RER.JD  category1
1  0.635806  1.210649  0.319687  INM.LE  category1
2  0.578831  0.103654  1.333646  BSZ.HS  category1
3 -1.128907 -1.189098  1.531494  GJZ.UX  category1
4  0.067668 -1.990996  0.088281  IQZ.KS  category1
 Copy code 

Let's draw 3D Bubble chart of , Since it is a three-dimensional figure, it shows that there is x Axis 、y Axis and z Axis , The code is as follows

cf.datagen.scatter3d(2,150).iplot(kind='scatter3d',x='x',y='y',z='z',size=15,
                                  categories='categories',text='text',
                                  title='Cufflinks - 3D Bubble chart ',colors=['yellow','purple'],
                                  width=1,margin=(0,0,0,0),
                                  opacity=1)
 Copy code 

output

3D Bubble chart

So it's mentioned that 3D Scatter plot , I have to say 3D Bubble chart of , Suppose our data set is this long

cf.datagen.bubble3d(5,4).head()
 Copy code 

output

          x         y         z  size    text categories
0 -1.888528  0.801430 -0.493671    77  OKC.HL  category1
1 -0.744953 -0.004398 -1.249949    61  GAG.UH  category1
2  0.980846  1.241730 -0.741482    37  LVB.EM  category1
3 -0.230157  0.427072  0.007010    78  NWZ.MG  category1
4  0.025272 -0.424051 -0.602937    76  JDW.AX  category2
 Copy code 

Let's draw 3D Bubble chart of

cf.datagen.bubble3d(5,4).iplot(kind='bubble3d',x='x',y='y',z='z',size='size',
                               text='text',categories='categories',
                               title='Cufflinks - 3D Bubble chart ',colorscale='set1',
                               width=.9,opacity=0.9)
 Copy code 

output

Box figure

Next, let's look at the drawing of the box diagram , The box chart is useful for us to observe the distribution of data 、 Whether there are extreme values is of great help

df.iplot(kind = "box")
 Copy code 

output

Heat map

This is the drawing of the heat map , Let's look at the dataset

cf.datagen.heatmap(20,20).head()
 Copy code 

output

           y_0        y_1        y_2  ...       y_17       y_18       y_19
x_0  40.000000  58.195525  55.355233  ...  77.318287  80.187609  78.959951
x_1  37.111934  25.068114  25.730511  ...  27.261941  32.303315  28.550340
x_2  54.881357  54.254479  59.434281  ...  75.894161  74.051203  72.896999
x_3  41.337221  39.319033  37.916613  ...  15.885289  29.404226  26.278611
x_4  42.862472  36.365226  37.959368  ...  24.998608  25.096598  32.413760
 Copy code 

Let's draw a heat map , The code is as follows

cf.datagen.heatmap(20,20).iplot(kind='heatmap',colorscale='spectral',title='Cufflinks -  Heat map ')
 Copy code 

output

Trend chart

The so-called trend chart , To put it bluntly, it is the combination of line chart and area chart , The code is as follows

df[["A", "B"]].iplot(kind = 'spread')
 Copy code 

output

The pie chart

Now let's take a look at the drawing of pie chart , The code is as follows

cf.datagen.pie(n_labels=6, mode = "stocks").iplot(
    kind = "pie",
    labels = "labels",
    values = "values")
 Copy code 

output

K Line graph

cufflinks It can also be used to draw K Line graph , Let's look at the data set here

cf.datagen.ohlc().head()
 Copy code 

output

                  open        high         low       close
2015-01-01  100.000000  119.144561   97.305961  106.125985
2015-01-02  106.131897  118.814224   96.740816  115.124342
2015-01-03  116.091647  131.477558  115.801048  126.913591
2015-01-04  128.589287  144.116844  117.837221  136.332657
2015-01-05  134.809052  138.681252  118.273850  120.252828
 Copy code 

As you can see from the data set above , There is an opening price 、 Closing price 、 The highest / The lowest price , Then let's draw K Line graph

cf.datagen.ohlc().iplot(kind = "ohlc",xTitle = " date ",
                        yTitle=" Price ",title = "K Line graph ")
 Copy code 

output

Histogram

df = pd.DataFrame({'a': np.random.randn(1000) + 1, 'b': np.random.randn(1000),
                    'c': np.random.randn(1000) - 1}, columns=['a', 'b', 'c'])
df.iplot(kind = "histogram")
 Copy code 

output

Drawing of multiple subgraphs

Then let's look at the drawing of multiple subgraphs , One is to use scatter_matrix() Method to implement

df = pd.DataFrame(np.random.randn(1000, 4),
                  columns=['a', 'b', 'c', 'd'])
df.scatter_matrix()
 Copy code 

output

The other is to use subplots Parameters , Set its parameters to True, For example, let's draw multiple histogram subgraphs

df_h=cf.datagen.histogram(4)
df_h.iplot(kind='histogram',subplots=True,bins=50)
 Copy code 

output

Or draw multiple line graph subgraphs

df=cf.datagen.lines(4)
df.iplot(subplots=True,subplot_titles=True,legend=True)
 Copy code 

output

Finally, we can freely combine the drawing of multiple subgraphs , Through the inside specs Parameters

df=cf.datagen.bubble(10,50,mode='stocks')
#  Define the form of the chart to be drawn 
figs=cf.figures(df,[dict(kind='histogram',keys='x',color='blue'),
                    dict(kind='scatter',mode='markers',x='x',y='y',size=5),
                    dict(kind='scatter',mode='markers',x='x',y='y',size=5,color='teal')],asList=True)
figs.append(cf.datagen.lines(1).figure(bestfit=True,colors=['blue'],bestfit_colors=['red']))
base_layout=cf.tools.get_base_layout(figs)
#  How multiple subgraphs are distributed ,specs Among the parameters , It is divided into two rows and two columns for distribution 
specs=cf.subplots(figs,shape=(3,2),base_layout=base_layout,vertical_spacing=.25,horizontal_spacing=.04,
               specs=[[{'rowspan':2},{}],[None,{}],[{'colspan':2},None]],
               subplot_titles=[' Histogram ',' Scatter plot _1',' Scatter plot _2',' Broken line diagram + Fitting line '])
specs['layout'].update(showlegend=True)
cf.iplot(specs)
 Copy code 

output