The easiest way! ! Use python to generate dynamic bar graphs (solve error reporting!)

The easiest way! ! Generate dynamic bar graph with python

Recently, dynamic bar graphs are very popular. On video sites such as station B, such videos often have millions of views. Today, we use a third-party library: bar_chart_race (version 0.2) to generate dynamic bar graphs; The generated effect is shown in the figure:
Problem: I found that following the operations in the current popular articles, there will be a lot of errors , some of which I have solved, and some of which I have not solved. Now record the current progress here; the data and procedures are in Here: click to download ;

1. Install the third-party package: bar_chart_race

First of all, we must understand: only version 0.2 supports dynamic bar graphs, so we must ensure that the third-party library we install is version 0.2;
unfortunately, the third-party package we installed through the pip install bar_chart_race command is version 0.1. It cannot be upgraded either; therefore, we recommend downloading on github and then installing; the installation steps are as follows:

1. First of all, make sure you have installed the github software. If not, please download it here for windows: https://git-scm.com/download/win , click on your own version to download, and then install; the same goes for other systems~~
   

2. After git is installed, switch to the directory you want to download in cmd, and download: git clone https://github.com/dexplo/bar_chart_race.git
   Among them, https://github.com/dexplo/bar_chart_race.git is its git address;
   the directory I downloaded here is:, the D:\PY_useful\3_python 基础download is complete After that, a folder will be generated:bar_chart_race
   

3. Enter the folder in cmd, and then run:, python setup.py installafter the installation is complete, it will prompt:, Finished processing dependencies for bar-chart-race==0.2.0seeing this, it proves that we are successful!
   

2. Data preparation

The data needs to meet the following conditions:

1. Each row must be a single date of data;
2. Each column is a separate category;
3. The index is preferably date data (optional);
4. The data must be changed to pandas.DataFrame format;

Official website https://www.dexplo.org/bar_chart_race/, examples of official websites are as follows:

date	Belgium	China	France	Germany	Iran	Italy	Netherlands	Spain	USA	United Kingdom
2020-04-08	2240	3337	10887	2349	3993	17669	2255	14792	14704	7111
2020-04-09	2523	3339	12228	2607	4110	18279	2403	15447	16553	7993
2020-04-10	3019	3340	13215	2767	4232	18849	2520	16081	18595	8974
2020-04-11	3346	3343	13851	2894	4357	19468	2653	16606	20471	9892
2020-04-12	3600	3343	14412	3022	4474	19899	2747	17209	22032	10629

However, the official website are examples of the use of df = bcr.load_dataset('covid19_tutorial')load, and if we are not over the wall, it is unable to load the data, the data I put on here for everyone to practice: Click to download training data ;

3. Function explanation

The third-party module bar_chart_race has only two functions to generate our dynamic graph,

1. Dynamic bar graph:; bar_chart_race.bar_chart_race()The following explains in detail all the parameters in this function:
2. Dynamic line graph bar_chart_race.line_chart_race():; similar to the bar graph above;

import bar_chart_race as bcr
import pandas
df=pd.read_csv('数据.csv',index_col=0)   # 下载数据后，放在.py文件所在文件夹；

# 以下有 # 注释的参数，是我实际使用中发现并没有的；可是官网资料上有这些参数，不知道为什么~~
bcr.bar_chart_race(
    df=df,    # 第一个参数就是数据，这个数据格式必须是 pandas.DataFrame 格式,同时满足数据准备中所说的条件；
    filename='bar_chart.mp4',   # 这个参数是生成文件的名字，一般为.mp4 & .gif；
    orientation='h',    # 方向
    sort='desc', # 排序
    n_bars=6, # 限制条形图数量
    fixed_order=False,  # 固定标签
    fixed_max=True, # 固定轴的最大值
    steps_per_period=10,   # 帧数设置
    interpolate_period=False,  # 插入时间
    # label_bars=True,   # 是否有label
    bar_size=.95,   # 设置bar宽度 取值 0~1 之间；
    period_label={
    
    'x': .99, 'y': .25, 'ha': 'right', 'va': 'center'},
    # period_fmt='%B %d, %Y',  # 日期的格式设置
    period_summary_func=lambda v, r: {
    
    'x': .99, 'y': .18,
                                      's': r'Total weigth: {v.sum():,.0f}',
                                      'ha': 'right', 'size': 8, 'family': 'Courier New'},
    perpendicular_bar_func='median',
    period_length=500,
    # figsize=(5, 3),
    # dpi=144,
    # cmap='dark12',
    title='COVID-19 Deaths by Country',
    # title_size='',
    # bar_label_size=7,
    # tick_label_size=7,
    shared_fontdict={
    
    'family' : 'DejaVu Sans', 'color' : '.1'},
    scale='linear',
    writer=None,
    fig=None,
    bar_kwargs={
    
    'alpha': .7},
    filter_column_colors=False) 

4. Detailed examples (copy and use)

4.1 Basic bar chart

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,filename='动态条形图.mp4')

The effect is as follows:

4.2 Histogram

Using orientationparameters of the conversion pattern; there orientationare two values may be assigned hand v;

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,'动态柱状图.mp4',orientation='v')

The effect is as follows:

4.3 Sorting diagram

By default, it is sorted in descending order. When we need to sort in ascending order, use the parameter:sort='asc'

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,'动态柱状图_升序.mp4',sort='asc')

The effect is as follows:

4.4 Limit the number of bars

By default, all columns will be displayed; if we need to control the number, use the parameter:, n_bars=6where the number 6 can be changed to any integer;

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,'动态柱状图_5条形图.mp4',n_bars=5)

The effect is as follows:

4.5 Order of fixed bars

By default, the bar graph will be sorted in descending order. If we want to fix the order, use the parameter:, fixed_orderchange it to True or assign a list;

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,'动态柱状图_条形固定.mp4',fixed_order=['Apple', 'banana', 'watermelon', 'orange'])

The effect is as follows:

4.6 Maximum fixed axis

By default, the maximum value according to the time axis conversion, if desired fixed parameters: fixed_max=True.

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,'动态条形图_固定轴的最大值.mp4',fixed_max=True)

The effect is as follows:

4.7 Set the number of frames

By default, the number of frames is 10, we can increase this value to make the image smoother; of course, it can also be reduced;

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,'动态条形图_设置帧数.mp4',steps_per_period=24)

The effect is as follows:

4.8 Setting the step length and dwell time

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,'动态条形图_设置步长与停留时间.mp4',steps_per_period=20, period_length=200)

The effect is as follows:

4.9 Add insertion time

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,'动态条形图_插入时间.mp4',interpolate_period=True)

4.10 Set font properties

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,filename='动态条形图_设置字体属性.mp4',shared_fontdict={
    
    'family': 'Helvetica', 'weight': 'bold','color': 'rebeccapurple'})

The effect is as follows:

4.11 Customize bar attributes

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,filename='动态条形图_定制条形属性.mp4',bar_kwargs={
    
    'alpha': .2, 'ec': 'black', 'lw': 3})

The effect is as follows:

4.12 Add text description

This, the research has not been successful yet, update later~

4.13 Increase the average vertical bar

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,filename='动态条形图_增加垂直bar平均值.mp4',perpendicular_bar_func='mean')

The effect is as follows:

4.14 Custom vertical bar-maximum

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
# 自定义增加垂直bar
def func(values, ranks):
    return values.max()
bcr.bar_chart_race(df,filename='动态条形图_增加垂直bar自定义最大值.mp4', perpendicular_bar_func=func)

The effect is as follows:

4.15 Bar graph colors are not repeated

import bar_chart_race as bcr
import pandas as pd

df=pd.read_csv('数据.csv',index_col=0)
bcr.bar_chart_race(df,filename='动态条形图_条形图颜色不重复.mp4',filter_column_colors=True)

The effect is as follows:

5. Problems encountered

5.1 [error] AttributeError: module ‘sip‘ has no attribute ‘setapi‘

Problem description: Import import bar_chart_race error report I
recently bar_chart_raceencountered an error when importing the package: AttributeError: module'sip' has no attribute'setapi'
Imported package:

import bar_chart_race as bcr

Just run the imported statement, it will report an error: AttributeError: module'sip' has no attribute'setapi'

Solution:
Finally, I found that the reason for this error was that the version of the matplotlib library was too high. The version I reported was 3.3.2. There is no problem if I downgrade it to 3.3.0;

Uninstall first:pip uninstall matplotlib

Install again:pip install matplotlib==3.3.0

problem solved! !

5.2 [Report Error] IndexError: list index out of range

Problem description: When the output is in .gif format (if it is in .mp4 format, no error will be reported), an error will be reported during runtime:IndexError: list index out of range

Solution: Download : download https://imagemagick.org/script/download.phpthe software at the site, just install it, but this URL is slow to download, you can also click here to download ( click to download ); Download and install, remember to check these two

items when installing ; after the installation is complete, the problem is solved ! !

5.3 [Report Error] TypeError: bar_chart_race() got an unexpected keyword argument 'figsize'

Problem description: When configuring according to the example on the official website, it is found that some parameters are not in the actual function, listed as follows:

    label_bars=True,   # 是否有label
    period_fmt='%B %d, %Y',  # 日期的格式设置
    figsize=(5, 3),
    dpi=144,
    cmap='dark12',
    title_size='',
    bar_label_size=7,
    tick_label_size=7,

The parameters listed above are not included in the function, and I don’t know why~ Because the article about this third-party package, everyone is an example of the official website of the translation, and no one actually verified it, so this question, if you have You know the reason~ Welcome everyone to tell me~ Thanks! !