#11 Pandas教学3

Pandas汇入CSV档

# 载入pandas
import pandas as pd

if __name__ == "__main__":
    #汇入CSV来建立DataFrame
    data = pd.read_csv("./car.csv")
    print(data)
    """
             brand   auto    price
    0    Toyota 86  False  1300000
    1   Subaru BRZ   True  1320000
    2  NISSAN GT-R   True  6750000
    3   MAZDA MX-5  False  1340000
    """

而且Pandas很聪明，会自己辨识资料栏位应该要是什麽资料型态

# 载入pandas
import pandas as pd

if __name__ == "__main__":
    #汇入CSV来建立DataFrame
    data = pd.read_csv("./car.csv")
    print(data)
    """
             brand   auto    price
    0    Toyota 86  False  1300000
    1   Subaru BRZ   True  1320000
    2  NISSAN GT-R   True  6750000
    3   MAZDA MX-5  False  1340000
    """

    print(data["brand"].dtype) #object
    print(data["auto"].dtype) #bool
    print(data["price"].dtype) #int64

指定某个栏位为索引

# 载入pandas
import pandas as pd

if __name__ == "__main__":
    #汇入CSV来建立DataFrame
    data = pd.read_csv("./car.csv", index_col = ["brand"])
    print(data)
    """
               auto    price
    brand
    Toyota 86    False  1300000
    Subaru BRZ    True  1320000
    NISSAN GT-R   True  6750000
    MAZDA MX-5   False  1340000
    """

指定某个栏位的资料型态

# 载入pandas
import pandas as pd
# 载入numpy
import numpy as np

if __name__ == "__main__":
    #汇入CSV来建立DataFrame
    data = pd.read_csv("./car.csv", dtype={"price":np.float64})
    print(data)
    """
         brand   auto      price
    0    Toyota 86  False  1300000.0
    1   Subaru BRZ   True  1320000.0
    2  NISSAN GT-R   True  6750000.0
    3   MAZDA MX-5  False  1340000.0
    """

    print(data["brand"].dtype) #object
    print(data["auto"].dtype) #bool
    print(data["price"].dtype) #float64  

自订栏位名称

# 载入pandas
import pandas as pd
# 载入numpy
import numpy as np

if __name__ == "__main__":
    #汇入CSV来建立DataFrame
    data = pd.read_csv("./car.csv", names = ['A', 'AA', 'AAA'])
    print(data)
    """
             A     AA      AAA
    0        brand   auto    price
    1    Toyota 86  FALSE  1300000
    2   Subaru BRZ   TRUE  1320000
    3  NISSAN GT-R   TRUE  6750000
    4   MAZDA MX-5  FALSE  1340000
    """

    print(data["AAA"])
    """
    0      price
    1    1300000
    2    1320000
    3    6750000
    4    1340000
    """

这时候会出现一个问题，原本的栏位名称变成第一笔资料了，必须解决他

# 载入pandas
import pandas as pd
# 载入numpy
import numpy as np

if __name__ == "__main__":
    #汇入CSV来建立DataFrame
    data = pd.read_csv("./car.csv", names = ['A', 'AA', 'AAA'], header=0)
    print(data)
    """
             A     AA      AAA
    0    Toyota 86  False  1300000
    1   Subaru BRZ   True  1320000
    2  NISSAN GT-R   True  6750000
    3   MAZDA MX-5  False  1340000
    """

    print(data["AAA"])
    """
    0      price
    1    1300000
    2    1320000
    3    6750000
    4    1340000
    """

输出为CSV档

# 载入pandas
import pandas as pd

if __name__ == "__main__":
    # 以字典来建立DataFrame，把brand栏位设为索引
    d = {
        "brand":["Toyota 86", "Subaru BRZ", "NISSAN GT-R", "MAZDA MX-5"],
        "auto":[False, True, True, False],
        "price":[1300000, 1320000, 6750000, 1340000],
        }
    data = pd.DataFrame(d)
    print(data)
    """
             brand   auto    price
    0    Toyota 86  False  1300000
    1   Subaru BRZ   True  1320000
    2  NISSAN GT-R   True  6750000
    3   MAZDA MX-5  False  1340000
    """

    #输出为CSV档
    data.to_csv("./new_car.c