1. 多个一维list,长度一样,每个代表一列,如何拼接成一个NumPy Array,然后转换成Pandas DataFrame?
import numpy as np
import Pandas as pd
# 比如下面这样三列数据:
col1 = [1,2,3,4,5]
col2 = ['a', 'b','c','d','e']
col3 = [99,105, 50, 80, 100]
col4 = [70, 90, 95, 80, 60]
# 拼接成NumPy Array:
data_np1 = np.array([col1, col2, col3]).T
data_np1
# 显示结果:
array([['1', 'a', '99'],
['2', 'b', '105'],
['3', 'c', '50'],
['4', 'd', '80'],
['5', 'e', '100']], dtype='<U21')
# 转换成Pandas DataFrame:
data_pd1 = pd.DataFrame(data_np1)
# 上面一行可以指定列名称,就像下面这样:
data_pd1 = pd.DataFrame(data_np1, columns=['index', 'name', 'score'])
2. Numpy Array和Pandas DataFrame之间的转换?
# NumPy Array转Pandas DataFrame
data_pd = pd.DataFrame(data_np)
data_pd = pd.DataFrame(data_np,columns=['col1',....])
# Pandas DataFrame转NumPy Array
data_np2 = data_pd.values
3. 多个NumPy Array如何拼接?
可以使用hstack、vstack实现水平和垂直方向的拼接,也可使用np.concatenate,通过axis实现拼接方向控制
data_np2 = np.array([col4]).T
# 水平方向上的拼接
data_np = np.hstack((data_np1, data_np2))
data_np = np.concatenate((data_np1, data_np2), axis=1)