大家好，我是欧K~
本期我们通过分析40000+汽车之家在售汽车信息数据，看看国民消费等级以及各品牌汽车性能(排列/油耗)情况等等，希望对小伙伴们有所帮助，如有疑问或者需要改进的地方可以私信小编。
涉及到的库：
- Pandas — 数据处理
- Pyecharts — 数据可视化
可视化部分：
- 柱状图 — Bar
- 折线图 — Line
- 饼图 — Pie
- 象形图 — PictorialBar
- 词云图 — stylecloud
- 词云图 — stylecloud
- 组合组件 — Grid
1. 导入模块
import stylecloud import pandas as pd from PIL import Image from pyecharts.charts import Bar from pyecharts.charts import Line from pyecharts.charts import Grid from pyecharts.charts import Pie from pyecharts.charts import PictorialBar from pyecharts import options as opts from pyecharts.globals import ThemeType
2. Pandas数据处理
2.1 读取数据
df = pd.read_csv('autohome.csv', encoding='gbk', low_memory=False) df.head(5)
结果：
2.2 数据大小
df.shape
(40460, 18），可以看到一共有: 40460条数据，包含18个字段。
2.3 筛选部分列数据
df1 = df.loc[:,['full_name', 'name', 'brand_name', 'group_name', 'series_name', 'price', 'year', 'carType', 'displacement', 'month', 'chexi', 'oil', 'chargetime', 'color']]
2.4 查看索引、数据类型和内存信息
df1.info()
2.5 查看所有汽车品牌
df1['brand_name'].unique()
一共有 237 个汽车品牌。
3. Pyecharts可视化
3.1 汽车售价区间占比饼图
price_bin = pd.cut(df_tmp['price'],bins=[0,10,30,50,70,100,500,7000],include_lowest=True,right=False, labels=['<10万', '10-30万', '30-50万', '50-70万', '70-100万', '100-500万', '>500万']) df_price = pd.value_counts(price_bin) data_pair = [list(z) for z in zip(df_price.index.tolist(), df_price.values.tolist())] p1 = ( Pie(init_opts=opts.InitOpts(theme=ThemeType.DARK,width='1000px',height='600px',bg_color='#0d0735')) .add( '售价', data_pair, radius=['40%', '70%'], label_opts=opts.LabelOpts( position="outside", formatter="{a|{a}}{abg|}\n{hr|}\n {b|{b}: }{c|{c}} {d|{d}%} ", background_color="#eee", border_color="#aaa", border_width=1, border_radius=4, rich={ "a": {"color": "#c92a2a", "lineHeight": 20, "align": "center"}, "abg": { "backgroundColor": "#00aee6", "width": "100%", "align": "right", "height": 22, "borderRadius": [4, 4, 0, 0], }, "hr": { "borderColor": "#00d1b2", "width": "100%", "borderWidth": 0.5, "height": 0, }, "b": {"color": "#bc0024","fontSize": 16, "lineHeight": 33}, "c": {"color": "#4c6ef5","fontSize": 16, "lineHeight": 33}, "d": {"color": "#bc0024","fontSize": 20, "lineHeight": 33}, }, ), itemstyle_opts={ 'normal': { 'shadowColor': 'rgba(0, 0, 0, .5)', 'shadowBlur': 5, 'shadowOffsetY': 2, 'shadowOffsetX': 2, 'borderColor': '#fff' } } ) .set_global_opts( title_opts=opts.TitleOpts( title="汽车售价区间占比", pos_left='center', pos_top='center', title_textstyle_opts=opts.TextStyleOpts( color='#ea1d5d', font_size=26, font_weight='bold' ), ), visualmap_opts=opts.VisualMapOpts( is_show=False, min_=0, max_=20000, is_piecewise=False, dimension=0, range_color=['#e7e1ef','#d4b9da','#c994c7','#df65b0','#e7298a','#ce1256','#91003f'] ), legend_opts=opts.LegendOpts(is_show=False), ) )
效果：
3.2 xx品牌汽车百公里油耗/排量/价格
colors = ["#ea1d5d", "#00ad45", "#0061d5"] width = 3 car_brand_name = '凯迪拉克' # car_brand_name = '玛莎拉蒂' data_tmp = df1[(df1.brand_name == car_brand_name)] data = data_tmp.copy() data = data.dropna(subset=['price','displacement','oil']) data = data[(data['price']>0) & (data['displacement']>0) & (data['oil']>0)] data['price'] = data['price'] /10000 price = data.price.values.tolist() displacement = data.displacement.values.tolist() oil = data.oil.tolist() region = [i for i in range(len(price))] ... line2.overlap(bar1) grid2 = Grid(init_opts=opts.InitOpts(width='1000px', height='600px',bg_color='#0d0735')) grid2.add(line2, opts.GridOpts(pos_top="20%",pos_left="5%", pos_right="15%"), is_control_axis_index=True)
效果：
3.3 在售汽车品牌数量TOP15象形图
df_brand_name_tmp = df1.groupby(['brand_name'])['name'].count().to_frame('数量').reset_index().sort_values(by=['数量'],ascending=False) df_brand_name = df_brand_name_tmp[:15] x_data = df_brand_name['brand_name'].values.tolist()[::-1] y_data = df_brand_name['数量'].values.tolist()[::-1] for idx,sch in enumerate(x_data): icons.append(dict(name=sch, value=y_data[idx], symbol=sch_icons[sch])) p1 = ( PictorialBar(init_opts=opts.InitOpts(theme='light', width='1000px', height='700px')) .add_xaxis(x_data) .add_yaxis('', icons, label_opts=opts.LabelOpts(is_show=False), category_gap='40%', symbol_repeat='fixed', symbol_margin='30%!', symbol_size=40, is_symbol_clip=True, itemstyle_opts={"normal": { 'shadowBlur': 10, 'shadowColor': 'rgba(0, 0, 200, 0.3)', 'shadowOffsetX': 10, 'shadowOffsetY': 10,} } ) .set_global_opts( title_opts=opts.TitleOpts(title='在售汽车品牌数量TOP15',pos_top='2%',pos_left = 'center', title_textstyle_opts=opts.TextStyleOpts(color="blue",font_size=30)), xaxis_opts=opts.AxisOpts( position='top', is_show=True, axistick_opts=opts.AxisTickOpts(is_show=True), axislabel_opts=opts.LabelOpts(font_size=20,color='#ed1941',font_weight=700,margin=12), splitline_opts=opts.SplitLineOpts(is_show=True, linestyle_opts=opts.LineStyleOpts(type_='dashed')), axisline_opts=opts.AxisLineOpts(is_show=False, linestyle_opts=opts.LineStyleOpts(width=2, color='#DB7093')) ), yaxis_opts=opts.AxisOpts( is_show=True, is_scale=True, axistick_opts=opts.AxisTickOpts(is_show=False), axislabel_opts=opts.LabelOpts(font_size=20,color='#ed1941',font_weight=700,margin=20), axisline_opts=opts.AxisLineOpts(is_show=False, linestyle_opts=opts.LineStyleOpts(width=2, color='#DB7093')) ), ) .reversal_axis() )
效果：
3.4 在售汽车品牌数量TOP15堆叠图
- 可以看到大众汽车在售量遥遥领先，奔驰、宝马等品牌紧随其后
3.5 汽车品牌词云
brand_name_list = [] for idx, value in enumerate(df_brand_name_tmp.brand_name.values.tolist()): brand_name_list += [value] * (df_brand_name_tmp.数量.values.tolist())[idx] pic_name = '词云.png' stylecloud.gen_stylecloud( text=' '.join(brand_name_list), font_path=r'STXINWEI.TTF', palette='cartocolors.qualitative.Bold_5', max_font_size=100, icon_name='fas fa-car-side', background_color='#0d0735', output_name=pic_name, ) Image.open(pic_name)
4. 项目在线运行地址
篇幅原因，部分代码未展示，在线运行地址：https://www.heywhale.com/mw/project/628da3cb400cf42ac1d064d3
END
以上就是本期为大家整理的全部内容了，赶快练习起来吧，喜欢的朋友可以
点赞、点在看
也可以分享让更多人知道
