python - 如何消除 "unsupported operand type(s) for -: ' str' 和 'str' "错误如下？

Question

当我运行以下代码时，

def difference(dataset, interval=1):
    diff = list()
    for i in range(interval, len(dataset)):
        value = dataset[i] - dataset[i - interval]
        diff.append(value)
    return numpy.array(diff)

series = read_csv('Houston_weather.csv', header=None)
X = series.values
days_in_year = 365
differenced = difference(X, days_in_year)

我收到以下错误，有办法消除它吗？我正在研究天气的时间序列。

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-70-a0a67d687691> in <module>
      4 X = series.values
      5 days_in_year = 365
----> 6 differenced = difference(X, days_in_year)

<ipython-input-64-b7e841922d29> in difference(dataset, interval)
      3     diff = list()
      4     for i in range(interval, len(dataset)):
----> 5         value = dataset[i] - dataset[i - interval]
      6         diff.append(value)
      7     return numpy.array(diff)

TypeError: unsupported operand type(s) for -: 'str' and 'str'

我的数据框是:

Answer 1

它只是说“您正在尝试在 2 个字符串之间使用 - 运算符”。您应该将它们转换为数字。

value = int(dataset[i]) - int(dataset[i -interval]) 如果它们是整数，可能会有所帮助。

value = float(dataset[i]) - float(dataset[i -interval]) 如果它们是 float ，可能会有所帮助。

此外，pandas 的文档说最好使用 to_numpy() 而不是 values。

Warning We recommend using DataFrame.to_numpy() instead.

https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.values.html