在这种情况下，SettingWithCopyWarning是什么？如何避免呢？

Question

我正在尝试通过运行以下代码将数据框中的所有NaN值替换为零，但它似乎不起作用，并且得到警告：

# X is some DataFrame

for row in range(X.shape[0]):
    for col in range(X.shape[1]):
        if np.isnan(X.iloc[row, col]) == True:
            X.iloc[row, col] = 0.

SettingWithCopyWarning: A value is trying to be set on a copy of a slice from a DataFrame