Python Day8

@浙大疏锦行 PythonDay8.

内容：

• 字典（字典对）

  • dict_test = dict(name='zhangsan', age=18)

• 标签编码（离散特征 && 有顺序）
• 连续特征的归一化和标准化处理

代码：

# Question 1 dic
dict = {'Alice': 100, 'Bob': 200, 'Charlie': 300}

# Question 2 label-Code and 连续变量编码
import pandas as pd
import numpy as np

# 独热编码
def one_hot(data, columns):
    data = pd.get_dummies(data=data, columns=columns)
    return data
# 标签编码
def label_encoder(data, columns):
    if len(columns)  == 0:  # 处理异常情况
        return data
    value_index = list(range(len(columns))) # [1 2 3 4 5 6]
    mapping = dict(zip(columns, value_index))
    data[columns] = data[columns].map(mapping)
    return data
# 归一化
def min_max(data, columns):
    if len(columns)  == 0:  # 处理异常情况
        return data
    for column in columns:
        column_data = data[column]
        min_val = column_data.min()
        max_val = column_data.max()
        column_data = (column_data - min_val) / (max_val - min_val)
        data[column] = column_data
    return data

data = pd.read_csv("./data/heart.csv")
discrete_columns = []
continuous_columns = []
for column in data.columns:
    if data[column].dtype == 'object':
        discrete_columns.append(column)
    else:
        continuous_columns.append(column)

data = min_max(data, continuous_columns) # 归一化
# data = one_hot(data, discrete_columns) # 独热编码
data = label_encoder(data, discrete_columns) # 标签编码

print(data.head())