R语言数据分析案例，基于python实现的保险客户价值分析（聚类分析）

# 导入包 import pandas as pd import plotly.graph_objs as go #* 定义类 class EDAnalysis: def __init__(self, data:pd.DataFrame =None, # type: ignore id_col: str=None, # type: ignore target: str = None, # type: ignore cate_list: list = None, # type: ignore num_list: list = None, # type: ignore ): self.data = data self.id_col = id_col self.target = target self.num_list = num_list self.cate_list = cate_list def draw_bar(self, col_name: str): bar_num = self.data[col_name].value_counts() #! 条形图 trace0 = go.Bar(x=bar_num.index.tolist(), y=bar_num.values.tolist(), # type: ignore text=bar_num.values.tolist(), # type: ignore textposition='auto', marker=dict(color=["blue", "red", "green", "indianred", "darkgrey"], opacity=0.5) ) data = [trace0] layout = go.Layout(title=f'Distribution_num of {col_name}', bargap=0.4, height=600, xaxis={'title': col_name}) fig = go.Figure(data=data, layout=layout) return fig def draw_pie(self, col_name: str): pie_num = self.data[col_name].value_counts() #! 饼图 trace1 = go.Pie(labels=pie_num.index.tolist(), values=pie_num.values.tolist(), # type: ignore hole=.5, marker=dict(line=dict(color='white', width=1.3)) ) data = [trace1] layout = go.Layout(title=f'Distribution_ratio of {col_name}', height=600) fig = go.Figure(data=data, layout=layout) return fig def draw_bar_stack_cat(self, col_name: str): #! 交叉表 cross_table = round(pd.crosstab(self.data[col_name], self.data[self.target], normalize='index') * 100, 2) #! 索引 index_cols = cross_table.columns.tolist() # type: ignore #! 轨迹列表 data = [] for i in index_cols: trace = go.Bar(x=cross_table[i].values.tolist(), # type: ignore y=cross_table.index.tolist(), # type: ignore name=str(i), orientation='h', marker={'opacity': 0.8} ) data.append(trace) #! 布局 layout = go.Layout(title=f'Relationship Between {cross_table.index.name} and {cross_table.columns.name}', # type: ignore bargap=0.4, barmode='stack', height=600, xaxis={'title': '百分比'}, yaxis={'title': col_name} ) #! 画布 fig = go.Figure(data=data, layout=layout) return fig def draw_histogram(self, col_name: str): trace = go.Histogram(x=self.data[col_name], histnorm='probability', opacity=0.8) data = [trace] layout = go.Layout(title=f'Histogram of {col_name}', height=600, xaxis={'title': col_name}) fig = go.Figure(data=data, layout=layout) return fig def draw_bar_stack_num(self, col_name: str, bins_num:int = 25): #! 交叉表 x_data = pd.cut(self.data[col_name], bins=bins_num) cross_table = round(pd.crosstab(x_data, self.data[self.target], normalize='index') * 100, 2) #! 索引 index_cols = cross_table.columns.tolist() #type:ignore #! 轨迹列表 data = [] for i in index_cols: trace = go.Bar(x=cross_table.index.astype('str').tolist(), #type:ignore y=cross_table[i].values.tolist(), #type:ignore name=str(i), orientation='v', marker={'opacity': 0.8}, ) data.append(trace) #! 布局 layout = go.Layout(title=f'Relationship Between {cross_table.index.name} and {cross_table.columns.name}', #type:ignore bargap=0, barmode='stack', height=600, xaxis={'title': col_name}, yaxis={'title': '百分比'} ) #! 画布 fig = go.Figure(data=data, layout=layout) return fig def draw_scatter_matrix(self): #! 目标 index_vals = self.data[self.target].astype('category').cat.codes dimension_list = [] for i in self.num_list: dimension_list.append(dict(label=i, values=self.data[i])) trace = go.Splom(dimensions=dimension_list, text=self.data[self.target], marker=dict(color=index_vals, showscale=False, line_color='white', line_width=0.5) ) data = [trace] layout = go.Layout(title='ScatterPlot Matrix Between numeric Attributes', height=600) fig = go.Figure(data=data, layout=layout) return fig