python机器学习使数据更鲜活的可视化工具Pandas_Alive

python机器学习使数据更鲜活的可视化工具Pandas_Alive
目录
  • 安装方法
  • 使用说明
  • 支持示例展示
    • 水平条形图
    • 垂直条形图比赛
    • 条形图
    • 饼图
    • 多边形地理空间
    • 多个图表
  • 总结

    数据动画可视化制作在日常工作中是非常实用的一项技能。目前支持动画可视化的库主要以Matplotlib-Animation为主,其特点为:配置复杂,保存动图容易报错。

    安装方法

    ?12pip install pandas_alive # 或者conda install pandas_alive -c conda-forge

    使用说明

    pandas_alive 的设计灵感来自 bar_chart_race,为方便快速进行动画可视化制作,在数据的格式上需要满足如下条件

    • 每行表示单个时间段
    • 每列包含特定类别的值
    • 索引包含时间组件(可选)

    支持示例展示

    水平条形图

    ?123import pandas_alivecovid_df = pandas_alive.load_dataset()covid_df.plot_animated(filename='examples/perpendicular-example.webp',perpendicular_bar_func='mean')

    垂直条形图比赛

    ?123import pandas_alivecovid_df = pandas_alive.load_dataset()covid_df.plot_animated(filename='examples/example-barv-chart.webp',orientation='v')

    条形图

    与时间与 x 轴一起显示的折线图类似

    ?123456import pandas_alivecovid_df = pandas_alive.load_dataset()covid_df.sum(axis=1).fillna(0).plot_animated(filename='examples/example-bar-chart.webp',kind='bar', period_label={'x':0.1,'y':0.9}, enable_progress_bar=True, steps_per_period=2, interpolate_period=True, period_length=200)

    饼图

    ?123import pandas_alivecovid_df = pandas_alive.load_dataset()covid_df.plot_animated(filename='examples/example-pie-chart.webp',kind="pie",rotatelabels=True,period_label={'x':0,'y':0})

    多边形地理空间图

    ?1234567import geopandasimport pandas_aliveimport contextilygdf = geopandas.read_file('data/italy-covid-region.gpkg')gdf.index = gdf.regiongdf = gdf.drop('region',axis=1)map_chart = gdf.plot_animated(filename='examples/example-geo-polygon-chart.webp',basemap_format={'source':contextily.providers.Stamen.Terrain})

    多个图表

    pandas_alive 支持单个可视化中的多个动画图表。

    示例1

    ?123456789101112import pandas_aliveurban_df = pandas_alive.load_dataset("urban_pop")animated_line_chart = ( urban_df.sum(axis=1) .pct_change() .fillna(method='bfill') .mul(100) .plot_animated(kind="line", title="Total % Change in Population",period_label=False,add_legend=False))animated_bar_chart = urban_df.plot_animated(n_visible=10,title='Top 10 Populous Countries',period_fmt="%Y")pandas_alive.animate_multiple_plots('examples/example-bar-and-line-urban-chart.webp',[animated_bar_chart,animated_line_chart], title='Urban Population 1977 - 2018', adjust_subplot_top=0.85, enable_progress_bar=True)

    示例2

    ?123456import pandas_alivecovid_df = pandas_alive.load_dataset()animated_line_chart = covid_df.diff().fillna(0).plot_animated(kind='line',period_label=False,add_legend=False)animated_bar_chart = covid_df.plot_animated(n_visible=10)pandas_alive.animate_multiple_plots('examples/example-bar-and-line-chart.webp',[animated_bar_chart,animated_line_chart], enable_progress_bar=True)

    示例3

    ?1234567891011121314151617181920212223242526272829303132333435363738import pandas_aliveimport pandas as pddata_raw = pd.read_csv( "https://raw.githubusercontent.com/owid/owid-datasets/master/datasets/Long%20run%20life%20expectancy%20-%20Gapminder%2C%20UN/Long%20run%20life%20expectancy%20-%20Gapminder%2C%20UN.csv")list_G7 = [ "Canada", "France", "Germany", "Italy", "Japan", "United Kingdom", "United States",]data_raw = data_raw.pivot( index="Year", columns="Entity", values="Life expectancy (Gapminder, UN)")data = pd.DataFrame()data["Year"] = data_raw.reset_index()["Year"]for country in list_G7: data[country] = data_raw[country].valuesdata = data.fillna(method="pad")data = data.fillna(0)data = data.set_index("Year").loc[1900:].reset_index()data["Year"] = pd.to_datetime(data.reset_index()["Year"].astype(str))data = data.set_index("Year")animated_bar_chart = data.plot_animated( period_fmt="%Y",perpendicular_bar_func="mean", period_length=200,fixed_max=True)animated_line_chart = data.plot_animated( kind="line", period_fmt="%Y", period_length=200,fixed_max=True)pandas_alive.animate_multiple_plots( "examples/life-expectancy.webp", plots=[animated_bar_chart, animated_line_chart], title="Life expectancy in G7 countries up to 2015", adjust_subplot_left=0.2, adjust_subplot_top=0.9, enable_progress_bar=True)

    示例4

    ?123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110import geopandasimport pandas as pdimport pandas_aliveimport contextilyimport matplotlib.pyplot as pltimport urllib.request, jsonwith urllib.request.urlopen( "https://data.nsw.gov.au/data/api/3/action/package_show?id=aefcde60-3b0c-4bc0-9af1-6fe652944ec2") as url: data = json.loads(url.read().decode())# Extract url to csv componentcovid_nsw_data_url = data["result"]["resources"][0]["url"]# Read csv from data API urlnsw_covid = pd.read_csv(covid_nsw_data_url)postcode_dataset = pd.read_csv("data/postcode-data.csv")# Prepare data from NSW health datasetnsw_covid = nsw_covid.fillna(9999)nsw_covid["postcode"] = nsw_covid["postcode"].astype(int)grouped_df = nsw_covid.groupby(["notification_date", "postcode"]).size()grouped_df = pd.DataFrame(grouped_df).unstack()grouped_df.columns = grouped_df.columns.droplevel().astype(str)grouped_df = grouped_df.fillna(0)grouped_df.index = pd.to_datetime(grouped_df.index)cases_df = grouped_df# Clean data in postcode dataset prior to matchinggrouped_df = grouped_df.Tpostcode_dataset = postcode_dataset[postcode_dataset['Longitude'].notna()]postcode_dataset = postcode_dataset[postcode_dataset['Longitude'] != 0]postcode_dataset = postcode_dataset[postcode_dataset['Latitude'].notna()]postcode_dataset = postcode_dataset[postcode_dataset['Latitude'] != 0]postcode_dataset['Postcode'] = postcode_dataset['Postcode'].astype(str) # Build GeoDataFrame from Lat Long dataset and make map chartgrouped_df['Longitude'] = grouped_df.index.map(postcode_dataset.set_index('Postcode')['Longitude'].to_dict())grouped_df['Latitude'] = grouped_df.index.map(postcode_dataset.set_index('Postcode')['Latitude'].to_dict())gdf = geopandas.GeoDataFrame( grouped_df, geometry=geopandas.points_from_xy(grouped_df.Longitude, grouped_df.Latitude),crs="EPSG:4326")gdf = gdf.dropna() # Prepare GeoDataFrame for writing to geopackagegdf = gdf.drop(['Longitude','Latitude'],axis=1)gdf.columns = gdf.columns.astype(str)gdf['postcode'] = gdf.indexgdf.to_file("data/nsw-covid19-cases-by-postcode.gpkg", layer='nsw-postcode-covid', driver="GPKG") # Prepare GeoDataFrame for plottinggdf.index = gdf.postcodegdf = gdf.drop('postcode',axis=1)gdf = gdf.to_crs("EPSG:3857") #Web Mercator map_chart = gdf.plot_animated(basemap_format={'source':contextily.providers.Stamen.Terrain},cmap='cool')cases_df.to_csv('data/nsw-covid-cases-by-postcode.csv') from datetime import datetime bar_chart = cases_df.sum(axis=1).plot_animated( kind='line', label_events={ 'Ruby Princess Disembark':datetime.strptime("19/03/2020", "%d/%m/%Y"), 'Lockdown':datetime.strptime("31/03/2020", "%d/%m/%Y") }, fill_under_line_color="blue", add_legend=False) map_chart.ax.set_title('Cases by Location')grouped_df = pd.read_csv('data/nsw-covid-cases-by-postcode.csv', index_col=0, parse_dates=[0])line_chart = ( grouped_df.sum(axis=1) .cumsum() .fillna(0) .plot_animated(kind="line", period_label=False, title="Cumulative Total Cases", add_legend=False))def current_total(values): total = values.sum() s = f'Total : {int(total)}' return {'x': .85, 'y': .2, 's': s, 'ha': 'right', 'size': 11}race_chart = grouped_df.cumsum().plot_animated( n_visible=5, title="Cases by Postcode", period_label=False,period_summary_func=current_total) import timetimestr = time.strftime("%d/%m/%Y")plots = [bar_chart, line_chart, map_chart, race_chart]from matplotlib import rcParamsrcParams.update({"figure.autolayout": False})# make sure figures are `Figure()` instancesfigs = plt.Figure()gs = figs.add_gridspec(2, 3, hspace=0.5)f3_ax1 = figs.add_subplot(gs[0, :])f3_ax1.set_title(bar_chart.title)bar_chart.ax = f3_ax1 f3_ax2 = figs.add_subplot(gs[1, 0])f3_ax2.set_title(line_chart.title)line_chart.ax = f3_ax2f3_ax3 = figs.add_subplot(gs[1, 1])f3_ax3.set_title(map_chart.title)map_chart.ax = f3_ax3f3_ax4 = figs.add_subplot(gs[1, 2])f3_ax4.set_title(race_chart.title)race_chart.ax = f3_ax4timestr = cases_df.index.max().strftime("%d/%m/%Y")figs.suptitle(f"NSW COVID-19 Confirmed Cases up to {timestr}")pandas_alive.animate_multiple_plots( 'examples/nsw-covid.webp', plots, figs, enable_progress_bar=True)

    总结

    Pandas_Alive 是一款非常好玩、实用的动画可视化制图工具,以上就是python机器学习使数据更鲜活的可视化工具Pandas_Alive的详细内容,更多关于python机器学习可视化工具Pandas_Alive的资料请关注脚本之家其它相关文章!

    您可能感兴趣的文章:
    • python解释模型库Shap实现机器学习模型输出可视化
    • python数据挖掘使用Evidently创建机器学习模型仪表板
    • 在Python文件中指定Python解释器的方法
    • python机器学习Github已达8.9Kstars模型解释器LIME
    免责声明:本网信息来自于互联网,目的在于传递更多信息,并不代表本网赞同其观点。其原创性以及文中陈述文字和内容未经本站证实,对本文以及其中全部或者部分内容、文字的真实性、完整性、及时性本站不作任何保证或承诺,并请自行核实相关内容。本站不承担此类作品侵权行为的直接责任及连带责任。如若本网有任何内容侵犯您的权益,请及时联系我们,本站将会在24小时内处理完毕。
    相关文章
    返回顶部