Advanced visualization artifact Plotly mapping heat maps

Public account: You and the cabin by: Peter Editor: Peter

Hello, I’m Peter

There have been a number of recent updates on Plotly mapping. Today’s article is based on the official website and practical cases to explain how to draw thermal maps under different requirements.

There are 3 ways to draw heat maps in Plotly: Heatmap, imshow, and Figure_factory (Plotly’s graphic factory function)

Official website study address:

• Plotly.com/python/heat…
• Plotly.com/python/imsh…
• Plotly.com/python/anno…

directory

The overall contents of this article:

Serialized article by Plotly

For a few Plotly updates, check out:

Import libraries

import pandas as pd
import numpy as np

import plotly_express as px
import plotly.graph_objects as go
import plotly.figure_factory as ff  # Graphics Factory
from plotly.subplots import make_subplots  # Draw a subgraph
Copy the code

Display imshow based on PX

The imshow in Plotly is mainly used to display graphic data, but can also be used to display thermal maps:

Display RGB graph data

# Data section
rgb = np.array([[[105.0.0], [0.255.0], [0.0.55]],
                [[0.250.0], [0.0.205], [255.0.0]]], 
               dtype=np.uint8)

# call px
fig = px.imshow(rgb)
fig.show()
Copy the code

Based on nested lists

fig = px.imshow([[1.20.30],
                 [20.1.60]])

fig.show()
Copy the code

Data based on a two-dimensional array

1. Display color bar and color

data = np.arange(20).reshape(4.5) How to use Numpy to generate data
fig = px.imshow(data)
fig.show()
Copy the code

2, do not display color bar + gray

data = np.arange(20).reshape(4.5)
fig = px.imshow(data,binary_string=True)  # parameter Settings
fig.show()
Copy the code

3, display gray + color bar

data = np.arange(20).reshape(4.5)
fig = px.imshow(data,color_continuous_scale="gray")  # parameter Settings
fig.show()
Copy the code

Based on the data in the graphics file

1. Basic graphics

Import data from skimage
from skimage import data  

# import specified image data
img = data.astronaut()

fig = px.imshow(img, 
                binary_format="jpeg", 
                binary_compression_level=0)
fig.show()
Copy the code

2, do not display the axis scale

Import data from skimage
from skimage import data  

# import specified image data
img = data.astronaut()

fig = px.imshow(img, 
                binary_format="jpeg", 
                binary_compression_level=0)

# Do not display color range and axis scale
fig.update_layout(coloraxis_showscale=False)
fig.update_xaxes(showticklabels=False)
fig.update_yaxes(showticklabels=False)

fig.show()
Copy the code

Custom axis calibration value

import plotly.express as px

data=[[1.25.30.50.100], 
      [20.10.60.80.30], 
      [13.60.31.5.20]]

fig = px.imshow(
    data,  # incoming data
    Axis label Settings
    labels=dict(x="Week", y="Time of Day", color="Productivity"),  
    # Axis calibration Settings
    x=['Monday'.'Tuesday'.'Wednesday'.'Thursday'.'Friday'],  
    y=['early'.'in'.'late'])# Axis position adjustment
fig.update_xaxes(side="top")

fig.show()
Copy the code

Show imShow based on go

The Graph_objects method supports two methods to draw images:

• Go. Image: Supports only multi-channel Image data
• Go.Heatmap: Supports image data in a single channel

Based on the go.Image method

import plotly.graph_objects as go

# 2*3*3 list
rgb = [[[200.0.0], [0.200.0], [0.0.255]],
       [[0.255.0], [0.0.205], [255.0.0]]]

fig = go.Figure(go.Image(z=rgb))

fig.show()
Copy the code

Based on the go. HeatMap method

import plotly.graph_objects as go

fig = go.Figure(data=go.Heatmap(
                    z=[[10.20.30],
                      [20.1.60],
                      [30.60.10]]))

fig.update_layout(width=800,height=500)

fig.show()
Copy the code

Missing value handling

import plotly.graph_objects as go

fig = go.Figure(data=go.Heatmap(
                   z=[[1.None.30.50.1], [20.1.60, np.nan, 30], [30.60.1, -10.20]],
                   x=['Monday'.'Tuesday'.'Wednesday'.'Thursday'.'Friday'],
                   y=['Morning'.'Afternoon'.'Evening'],
                   hoverongaps = False))  # Missing value processing parameter
fig.show()
Copy the code

Based on the graph factory figure_Factory

An important function of figure_factory is to draw thermal maps with labels. We can see that all the above graphics are unmarked

Basic shapes

How to draw basic graphics to be marked based on graphic foreman:

import plotly.figure_factory as ff

z = [[1.1.3... 5.7..9.],
     [1.8..6..4..2.],
     [2..0.. 5.7..9.],
     [9..8..4..2..0]]

fig = ff.create_annotated_heatmap(z)
fig.show()
Copy the code

Custom colors

import plotly.figure_factory as ff

z = [[1.1.3... 5.7..9.],
     [1.8..6..4..2.],
     [2..0.. 5.7..9.],
     [9..8..4..2..0]]

# Add parameter: colorScale
fig = ff.create_annotated_heatmap(z,colorscale='Viridis')

fig.show()
Copy the code

How to set different colors:

import plotly.figure_factory as ff

z = [[1..3... 5.7.],
     [1.0.8..6..4.],
     [6..4..2..0.0],
     [9..7... 5.3.]]

This is more flexible by using nested lists to specify gradients for colors
colorscale = [[0.0.'green'], [0.5.'blue'], [1.'yellow']]

# specify the font color
font_colors = ['white'.'red']
Pass in the font and color parameters
fig = ff.create_annotated_heatmap(z, colorscale=colorscale, font_colors=font_colors)
fig.show()
Copy the code

Custom XY axis

import plotly.figure_factory as ff

# Display data
z = [[101.131.95],
     [112.74.146]]

# Two axes, optionally specified
x = ["Chinese"."Mathematics"."English"]
y = ["Xiao Ming"."Little red"]

The data to be annotated is generally the same as z
z_text = [[101.131.95],
          [112.74.146]]

fig = ff.create_annotated_heatmap(
    z,
    x=x,
    y=y,
    annotation_text=z_text, # Annotate text content
    colorscale='Viridis')

fig.show()
Copy the code

Set font size

Sometimes we change the font size in standard text:

import plotly.figure_factory as ff

# Display data
z = [[101.131.95],
     [112.74.146]]

# Two axes, optionally specified
x = ["Chinese"."Mathematics"."English"]
y = ["Xiao Ming"."Little red"]

The data to be annotated is generally the same as z
z_text = [[101.131.95],
          [112.74.146]]

fig = ff.create_annotated_heatmap(
    z,
    x=x,
    y=y,
    annotation_text=z_text, # Annotate text content
    colorscale='Viridis')

# font size Settings
for i in range(len(fig.layout.annotations)):
    fig.layout.annotations[i].font.size=20

fig.show()
Copy the code

Actual combat 1: Thermal map of correlation coefficient

Using stock data

Plotly_express built-in stock data; The correlation coefficient between two companies is solved by correlation coefficient function CORr function

stock = px.data.stocks()
stock.head()
Copy the code

A piece of data about a stock built into Plotly:

The correlation coefficient

Generate the correlation coefficient matrix while preserving two decimal places:

Coordinate axes and plot data

Generate axes and data to be plotted:

drawing

import plotly.figure_factory as ff

# Display data
z = data1

# Two axes, optionally specified
x = index
y = columns
# Display text content
z_text = data1

fig = ff.create_annotated_heatmap(
    z,  
    x=x,
    y=y,
    annotation_text=z_text, # Annotate text content
    colorscale='Viridis',
    showscale=True
)

# font size Settings
for i in range(len(fig.layout.annotations)):
    fig.layout.annotations[i].font.size=12

fig.show()
Copy the code

Case 2: Draw a thermal map of any coordinate axis

This case is a solution to a problem I once encountered, that is, how to solve the problem when there are two types of value and string in our coordinate axis.

Similar cases on official website

Both axes specify string data

import plotly.figure_factory as ff

z = [[101.131.95],
     [112.74.146]]

# Important: both are strings
x = ['Team A'.'Team B'.'Team C']
y = ['Game Two'.'Game One']

z_text = [[101.131.95],
          [112.74.146]]

fig = ff.create_annotated_heatmap(
    z,
    x=x,
    y=y,
    annotation_text=z_text, # Annotate text content
    colorscale='Viridis')

fig.show()
Copy the code

Here is a diagram:

Change the axis data

In the data of the coordinate axes, use both numbers and strings:

There is no picture: what is the problem?

How to solve it? Two methods are introduced:

Method 1: Set parameters

import plotly.figure_factory as ff

z = [[101.131.95],
     [112.74.146]]

# 1. Set the axis data and unify the data format
x = list(range(0.3))
y = list(range(0.2))

z_text = [[101.131.95],
          [112.74.146]]

fig = ff.create_annotated_heatmap(
    z,
    x=x,
    y=y,
    annotation_text=z_text, 
    colorscale='Viridis')

# Solutions
# Actual axis
xticks=[1.'Team B'.'Team C']  
yticks=[1.'Game One']  

# Change the coordinate axes: replace x with xticks, y is similar
fig.update_xaxes(tickvals=x, ticktext=xticks)
fig.update_yaxes(tickvals=y, ticktext=yticks)
fig.show()
Copy the code

Method 2: Customize the function

import plotly.figure_factory as ff

z = [[101.131.95],
     [112.74.146]]

x = [1.'Team B'.'Team C']
y = [1.'Game One']

# Data to display
z_text = z

# Key function: Modify the axis
def hm_axis(l) :
    return [f"{chr(0) +str(i)+chr(0)}" for i in l]

fig = ff.create_annotated_heatmap(
  z, 
  x=hm_axis(x), 
  y=hm_axis(y), 
  annotation_text=z_text, 
  colorscale='Viridis')
fig.show()
Copy the code

Perfect solution!