Pythonic方式处理索引-0

Question

我有来自Keras网的以下输出：

batch_cut_train = training_set_size - Y_train.shape[0]
batch_cut_test = abs(training_set_size - Y.shape[0]) - Y_test.shape[0]

first_output = model.predict(X_test, batch_size=batch_size)

output = pd.DataFrame(first_output, index=features[-Y_test.shape[0]-batch_cut_test:-batch_cut_test].index, columns=labels.columns)

当我的数据完全被batch_size整除时，我得到batch_cut_train = 0.然后，当我将其索引到要素中时，我得到了

features[:-0]

会抛出错误。什么是避免这种情况的干净方法？

错误为ValueError: Shape of passed values is (6, 1696), indices imply (6, 0)，因为它索引为0，因此功能（pd.DataFrame）中显示的范围基本上是空的。

编辑：这解决了它，但可能不是最好的方法：

batch_cut_train = training_set_size - Y_train.shape[0]
batch_cut_test = abs(training_set_size - Y.shape[0]) - Y_test.shape[0]


first_output = model.predict(X_test, batch_size=batch_size)

if batch_cut_test != 0:
    output = pd.DataFrame(first_output, index=features[-Y_test.shape[0]-batch_cut_test:-batch_cut_test].index, columns=labels.columns)
else:
    output = pd.DataFrame(first_output, index=features[-Y_test.shape[0]:].index, columns=labels.columns)

o = output.join(df['PX LAST'])
o.tail()