magigo / data_science_tool_book_code

9 stars 5 forks source link

KNN近邻算法中对鸢尾花的数据预处理似乎有问题? #5

Open rayepeng opened 5 years ago

rayepeng commented 5 years ago
def get_data(loc='iris.csv'):
    with open(loc, 'r') as fr:
        lines = csv.reader(fr)
        data_file = np.array(list(lines))
    data = data_file[1:, 0:-1].astype(float)
    labels = data_file[1:, -1]
    return data, labels

这里的切片我修改为了从0开始

还有一处地方

def try_once():
    data, labels = get_data()
    index = range(len(data))
    data = data[index]
    labels = labels[index]
    index = list(index)
    random.shuffle(index)
    labels = labels[index]
    data = data[index]
    input_data = data[-1]
    data = data[:-1]

python3中不能对rang类型的使用shuffle方法,所以我修改为先转换为list类型的