如何优化ABAP性能

       1、使用where语句

  不推荐

  Select * from zflight.

  Check : zflight-airln = ‘LF’ and zflight-fligh = ‘BW222’.

  Endselect.

  推荐

  Select * from zflight where airln = ‘LF’ and fligh = ‘222’.

  Endselect.

  2、使用聚合函数

  不推荐

  Maxnu = 0.

  Select * from zflight where airln = ‘LF’ and cntry = ‘IN’.

  Check zflight-fligh > maxnu.

  Maxnu = zflight-fligh.

  Endselect.

  推荐

  Select max( fligh ) from zflight into maxnu where airln = ‘LF’andcntry = ‘IN’.

  3、使用视图代替基本表查询

  不推荐

  Select * from zcntry where cntry like ‘IN%’.

  Select single * from zflight where cntry = zcntry-cntry andairln= ‘LF’.

  Endselect.

  推荐

  Select * from zcnfl where cntry like ‘IN%’ and airln = ‘LF’.

  Endselect.

  4、使用INTO table 代替select endselect

  不推荐

  Refresh: int_fligh.

  Select * from zflight into int_fligh.

  Append int_fligh. Clear int_fligh.

  Endselect.

  推荐

  Refresh: int_fligh.

  Select * from zflight into table int_fligh.

  5、使用批量修改内表代替逐行修改

  不推荐

  Loop at int_fligh.

  If int_fligh-flag is initial.

  Int_fligh-flag = ‘X’.

  Endif.

  Modify int_fligh.

  Endloop.

  推荐

  Int_fligh-flag = ‘X’.

  Modify int_fligh transporting flag where flag is initial.

  6、使用二分法查询,提高查询内表数据速度

  不推荐

  Read table int_fligh with key airln = ‘LF’.

  推荐

  Read table int_fligh with key airln = ‘LF’ binary search.

  7、两个内表添加使用批量增加代替逐行

  不推荐

  Loop at int_fligh1.

  Append int_fligh1 to int_fligh2.

  Endloop.

  推荐

  Append lines of int_fligh1 to int_fligh2.

  8、使用table buffering

  Use of buffered tables is recommended to improve theperformanceconsiderably. The buffer is bypassed while using thefollowingstatementsSelect distinct

  Select … for update

  Order by, group by, having clause

  Joins

  Use the Bypass buffer addition to the select clause in ordertoexplicitly bypass the buffer while selecting the data.

  9、 使用FOR ALL Entries

  不推荐

  Loop at int_cntry. Select single * from zfligh intoint_flighwhere cntry = int_cntry-cntry. Append int_fligh.Endloop.

  推荐

  Select * from zfligh appending table int_fligh

  For all entries in int_cntry

  Where cntry = int_cntry-cntry.

  10、正确地使用where语句,使查询能使用索引

  When a base table has multiple indices, the where clause shouldbein the order of the index, either a primary or a secondaryindex

  To choose an index, the optimizer checks the field namesspecifiedin the where clause and then uses an index that has thesame orderof the fields. One more tip is that if a table beginswith MANDT,while an index does not, there is a high possibilitythat theoptimizer might not use that index.

  11、正确地使用MOVE语句

  Instead of using the move-corresponding clause it is advisabletouse the move statement instead. Attempt should be made tomoveentire internal table headers in a single shot, rather thanmovingthe fields one by one.

  12、正确地使用inner join

  Let us take an example of 2 tables, zairln and zflight. Thetablezairln has the field airln, which is the airline code and thefieldlnnam, which is the name of the airline. The table zflight hasthefield airln, the airline code and other fields which holdthedetails of the flights that an airline operates.

  Since these 2 tables a re logically joined by the airln field,itis advisable to use the inner join.

  Select a~airln a~lnnam b~fligh b~cntry into table int_airdet

  From zairln as a inner join zflight as b on a~airln =b~airln.

  In order to restrict the data as per the selection criteria,awhere clause can be added to the above inner join.

  13、使用sort by 代替order by

  14、避免使用SELECT DISTINCT语句

  使用的 ABAP SORT + DELETE ADJACENT DUPLICATES 代替.

深度学习是机器学习的一个子领域,它基于人工神经网络的研究,特别是利用多层次的神经网络来进行学习和模式识别。深度学习模型能够学习数据的高层次特征,这些特征对于图像和语音识别、自然语言处理、医学图像分析等应用至关重要。以下是深度学习的一些关键概念和组成部分: 1. **神经网络(Neural Networks)**:深度学习的基础是人工神经网络,它是由多个层组成的网络结构,包括输入层、隐藏层和输出层。每个层由多个神经元组成,神经元之间通过权重连接。 2. **前馈神经网络(Feedforward Neural Networks)**:这是最常见的神经网络类型,信息从输入层流向隐藏层,最终到达输出层。 3. **卷积神经网络(Convolutional Neural Networks, CNNs)**:这种网络特别适合处理具有网格结构的数据,如图像。它们使用卷积层来提取图像的特征。 4. **循环神经网络(Recurrent Neural Networks, RNNs)**:这种网络能够处理序列数据,如时间序列或自然语言,因为它们具有记忆功能,能够捕捉数据中的时间依赖性。 5. **长短期记忆网络(Long Short-Term Memory, LSTM)**:LSTM 是一种特殊的 RNN,它能够学习长期依赖关系,非常适合复杂的序列预测任务。 6. **生成对抗网络(Generative Adversarial Networks, GANs)**:由两个网络组成,一个生成器和一个判别器,它们相互竞争,生成器生成数据,判别器评估数据的真实性。 7. **深度学习框架**:如 TensorFlow、Keras、PyTorch 等,这些框架提供了构建、训练和部署深度学习模型的工具和库。 8. **激活函数(Activation Functions)**:如 ReLU、Sigmoid、Tanh 等,它们在神经网络中用于添加非线性,使得网络能够学习复杂的函数。 9. **损失函数(Loss Functions)**:用于评估模型的预测与真实值之间的差异,常见的损失函数包括均方误差(MSE)、交叉熵(Cross-Entropy)等。 10. **优化算法(Optimization Algorithms)**:如梯度下降(Gradient Descent)、随机梯度下降(SGD)、Adam 等,用于更新网络权重,以最小化损失函数。 11. **正则化(Regularization)**:技术如 Dropout、L1/L2 正则化等,用于防止模型过拟合。 12. **迁移学习(Transfer Learning)**:利用在一个任务上训练好的模型来提高另一个相关任务的性能。 深度学习在许多领域都取得了显著的成就,但它也面临着一些挑战,如对大量数据的依赖、模型的解释性差、计算资源消耗大等。研究人员正在不断探索新的方法来解决这些问题。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

ChampaignWolf

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值