import json
from pyecharts import options as opts
from pyecharts.charts import Pie, Bar, Funnel, Page, PictorialBar, WordCloud,Line
from pyecharts.globals import ThemeType, SymbolType
from pyspark import SparkConf, SparkContext
import os
# 设置 Python 环境
os.environ["PYSPARK_PYTHON"] = r"C:\Users\h\AppData\Local\Programs\Python\Python38\python.exe"
# 配置 Spark
conf = SparkConf().setMaster("local[*]").setAppName("GameLevelDistribution")
sc = SparkContext(conf=conf)
# 导入 JSON 数据
with open("game_data.json", encoding="utf-8") as f:
data = json.load(f)
# 将数据转换为 RDD
rdd = sc.parallelize(data)
# 提取 level 数据
level_rdd = rdd.map(lambda x: x.get('level', '未知等级'))
# 统计每个 level 的出现次数
level_counts = level_rdd.countByValue()
# 将统计结果转换为列表，格式为 [("等级1", 次数1), ("等级2", 次数2), ...]
level_data = [(level, count) for level, count in level_counts.items()]
# 创建饼图
def drawPie():
# 按等级从小到大排序
sorted_level_data = sorted(level_data, key=lambda x: int(x[0].replace('级', '')))
c = (
Pie(init_opts=opts.InitOpts(theme=ThemeType.LIGHT))
.add("", sorted_level_data, center=["50%", "63%"])
.set_global_opts(
title_opts=opts.TitleOpts(
title="游戏等级分布",
title_textstyle_opts=opts.TextStyleOpts(color="white")
),
legend_opts=opts.LegendOpts(
pos_left="30%",
textstyle_opts=opts.TextStyleOpts(color="white"),
),
toolbox_opts=opts.ToolboxOpts(is_show=False), # 隐藏工具栏
datazoom_opts=None # 禁用数据缩放
)
.set_series_opts(
label_opts=opts.LabelOpts(formatter="{b}: {c}", color="white"),
itemstyle_opts=opts.ItemStyleOpts(border_width=0),
init_opts=opts.InitOpts(renderer='svg')
)
)
return c
# 统计游戏类型分布
word_counts = (
rdd.flatMap(lambda x: [word.strip() for word in x["type"].split("/") if word.strip()])
.map(lambda word: (word, 1))
.reduceByKey(lambda x, y: x + y)
.collect()
)
# 将统计结果转换为漏斗图需要的格式
funnel_data = [(word, count) for word, count in word_counts]
# 创建漏斗图
def drawFunnel():
l = (
Funnel(init_opts=opts.InitOpts(theme=ThemeType.LIGHT))
.add(
series_name="游戏类型分布",
data_pair=funnel_data,
gap=2,
tooltip_opts=opts.TooltipOpts(trigger="item", formatter="{a} <br/>{b} : {c}"),
label_opts=opts.LabelOpts(
is_show=True,
position="inside",
color="white"
),
itemstyle_opts=opts.ItemStyleOpts(
border_color="#fff",
border_width=0
),
)
.set_global_opts(
title_opts=opts.TitleOpts(
title="游戏类型分布漏斗图",
title_textstyle_opts=opts.TextStyleOpts(color="white")
),
legend_opts=opts.LegendOpts(
pos_top="bottom",
pos_left="center",
textstyle_opts=opts.TextStyleOpts(color="white")
),
toolbox_opts=opts.ToolboxOpts(is_show=False), # 隐藏工具栏
datazoom_opts=None # 禁用数据缩放
)
)
return l
# 统计游戏类型分布及平均评分
def split_types(game):
types = game["type"].split(" / ")
return [(t.strip(), 1) for t in types]
type_counts = rdd.flatMap(split_types).reduceByKey(lambda a, b: a + b).collectAsMap()
def split_types_with_score(game):
types = game["type"].split(" / ")
score = float(game["score"])
return [(t.strip(), score) for t in types]
type_scores = (
rdd.flatMap(split_types_with_score)
.groupByKey()
.mapValues(lambda scores: round(sum(scores) / len(scores), 2))
.collectAsMap()
)
# 准备数据
types = list(type_counts.keys())
counts = [type_counts[t] for t in types]
scores = [f"{score:.2f}" for score in [type_scores[t] for t in types]]
# 创建柱状图
def drawLine():
line = (
Line(init_opts=opts.InitOpts(theme="light"))
.add_xaxis(xaxis_data=types)
.add_yaxis(
series_name="游戏数量",
y_axis=counts,
markpoint_opts=opts.MarkPointOpts(
data=[
opts.MarkPointItem(type_="max", name="最大值"),
opts.MarkPointItem(type_="min", name="最小值"),
]
),
markline_opts=opts.MarkLineOpts(
data=[opts.MarkLineItem(type_="average", name="平均值")]
),
)
.add_yaxis(
series_name="平均评分",
y_axis=scores,
markpoint_opts=opts.MarkPointOpts(
data=[opts.MarkPointItem(type_="max", name="最高评分")]
),
markline_opts=opts.MarkLineOpts(
data=[opts.MarkLineItem(type_="average", name="平均评分")]
),
)
.set_global_opts(
title_opts=opts.TitleOpts(
title="游戏类型分布",
subtitle="纯属虚构",
title_textstyle_opts=opts.TextStyleOpts(color="white")
),
tooltip_opts=opts.TooltipOpts(trigger="axis", axis_pointer_type="cross"),
toolbox_opts=opts.ToolboxOpts(is_show=True),
xaxis_opts=opts.AxisOpts(type_="category", boundary_gap=False),
yaxis_opts=opts.AxisOpts(
name="数量/评分",
axislabel_opts=opts.LabelOpts(
formatter="{value}",
color="white"
)
),
legend_opts=opts.LegendOpts(textstyle_opts=opts.TextStyleOpts(color="white")),
datazoom_opts=None
)
)
return line
# 统计每一年的游戏数量
year_counts = rdd.map(lambda x: (x['year'], 1)).reduceByKey(lambda a, b: a + b).collect()
# 将结果转换为字典，方便后续处理
year_counts_dict = dict(year_counts)
# 分离“公元前”和普通年份
bc_years = {year: count for year, count in year_counts_dict.items() if '公元前' in year}
ad_years = {year: count for year, count in year_counts_dict.items() if '公元前' not in year}
# 对“公元前”年份进行排序（按绝对值降序）
sorted_bc_years = sorted(bc_years.keys(), key=lambda x: int(x.replace('公元前', '')), reverse=True)
# 对普通年份进行排序（按升序）
sorted_ad_years = sorted(ad_years.keys(), key=lambda x: int(x))
# 合并排序后的年份
sorted_years = sorted_bc_years + sorted_ad_years
# 准备柱状图数据
x_axis = sorted_years # 按排序后的年份
y_axis = [year_counts_dict[year] for year in sorted_years] # 对应的游戏数量
# 创建每年游戏数量统计的柱状图
def drawYearBar():
bar = (
Bar(init_opts=opts.InitOpts(theme=ThemeType.LIGHT))
.add_xaxis(x_axis)
.add_yaxis("游戏数量", y_axis)
.set_global_opts(
title_opts=opts.TitleOpts(
title="每年游戏数量统计",
title_textstyle_opts=opts.TextStyleOpts(color="white")
),
datazoom_opts=[opts.DataZoomOpts(), opts.DataZoomOpts(type_="inside")],
xaxis_opts=opts.AxisOpts(
axislabel_opts=opts.LabelOpts(
rotate=-15,
color="white"
)
),
yaxis_opts=opts.AxisOpts(
axislabel_opts=opts.LabelOpts(color="white")
),
legend_opts=opts.LegendOpts(textstyle_opts=opts.TextStyleOpts(color="white")),
toolbox_opts=opts.ToolboxOpts(is_show=False), # 隐藏工具栏
)
)
return bar
# 1. 桌游类型词云图
wordCount = (
rdd.flatMap(lambda x: [word.strip() for word in x["type"].split("/") if word.strip()])
.map(lambda word: (word, 1))
.reduceByKey(lambda x, y: x + y)
)
def draw_wordcloud() -> WordCloud:
wc = (
WordCloud()
.add(series_name="", data_pair=wordCount.collect(), word_size_range=[10, 60])
.set_global_opts(
title_opts=opts.TitleOpts(
title="桌游类型统计",
title_textstyle_opts=opts.TextStyleOpts(font_size=20, color="white"),
pos_left="center"
),
tooltip_opts=opts.TooltipOpts(is_show=True),
toolbox_opts=opts.ToolboxOpts(is_show=False), # 隐藏工具栏
datazoom_opts=None
)
)
return wc
# 2. 桌游时长象形柱状图（第11-20名）
selected_data = data[10:20]
titles = [game["title"] for game in selected_data]
times = [int(game["time_per_person"].replace("分/人", "")) for game in selected_data]
def draw_pictorialbar() -> PictorialBar:
pb = (
PictorialBar()
.add_xaxis(titles)
.add_yaxis(
"人均游戏时长（分钟）",
times,
label_opts=opts.LabelOpts(is_show=True, color="white"),
symbol_size=22,
symbol_repeat="fixed",
symbol_offset=[0, 0],
is_symbol_clip=True,
symbol=SymbolType.ROUND_RECT,
color="#37A2DA"
)
.reversal_axis()
.set_global_opts(
title_opts=opts.TitleOpts(
title="11-20名桌游人均耗时象形图",
title_textstyle_opts=opts.TextStyleOpts(color="white")
),
legend_opts=opts.LegendOpts(pos_top="5%"),
yaxis_opts=opts.AxisOpts(
axistick_opts=opts.AxisTickOpts(is_show=False),
axisline_opts=opts.AxisLineOpts(
linestyle_opts=opts.LineStyleOpts(opacity=0)
),
axislabel_opts=opts.LabelOpts(color="white"),
),
xaxis_opts=opts.AxisOpts(
name="时长（分钟）",
axislabel_opts=opts.LabelOpts(font_size=12, color="white"),
splitline_opts=opts.SplitLineOpts(is_show=True),
),
tooltip_opts=opts.TooltipOpts(is_show=True),
toolbox_opts=opts.ToolboxOpts(is_show=False), # 隐藏工具栏
datazoom_opts=None
)
)
return pb# 创建Page对象
page = Page(layout=Page.DraggablePageLayout)
# 添加图表
page.add(
drawPie(),
drawFunnel(),
drawLine(),
drawYearBar(),
draw_wordcloud(),
draw_pictorialbar()
)
# 将自定义HTML内容嵌入到最终的HTML文件中
html_content += page.render_embed()
html_content += "</body></html>"
# 将完整的HTML内容写入文件
with open("jishi2.html", "w", encoding="utf-8") as f:
f.write(html_content)
# 额外的逻辑，用于调整图表的排版
from bs4 import BeautifulSoup
# 打开并读取 HTML 文件
with open("jishi2.html", "r+", encoding='utf-8') as html:
html_bf = BeautifulSoup(html, 'lxml')
# 选择所有具有类名 'chart-container' 的 div 元素
divs = html_bf.select('.chart-container')
# 检查是否有足够的 div 元素
if len(divs) < 6:
print("HTML 文件中 .chart-container 元素不足 6 个")
else:
# 设置每个 div 的样式
divs[0]["style"] = "width:20%;height:30%;position:absolute;bottom:0;left:5%;" # 左下方
divs[1]["style"] = "width:25%;height:26%;position:absolute;top:20%;right:0;" # 右上方
divs[2]["style"] = "width:40%;height:30%;position:absolute;bottom:0;left:30%;" # 中间的下面
divs[3]["style"] = "width:25%;height:25%;position:absolute;top:50%;right:0;" # 右边的中间
divs[4]["style"] = "width:20%;height:10%;position:absolute;bottom:5%;right:3%;" # 右下角
divs[5]["style"] = "width:35%;height:40%;position:absolute;top:20%;left:1%;" # 左上方
# 修改 body 的样式
body = html_bf.find("body")
body["style"] = "background-image: url('img/backGround.png'); background-size: cover; background-repeat: no-repeat; background-attachment: fixed; background-color: #464646;"
# 将修改后的内容写回文件
html_new = str(html_bf)
html.seek(0, 0)
html.truncate()
html.write(html_new)
html.close()为什么这段代码运行出来，图表在页面中，鼠标可以改变图表的位置，我应该怎么解决这一问题
最新发布
本文介绍了一种在树形结构中计算删除节点及其子树步数期望值的方法。通过计算每个节点被选中的概率，利用这些概率求和得到最终的期望值。文中提供了一个具体的Python代码实现。
扫一扫
1941
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
