AI 助理
备案控制台
文档
输入文档关键字查找
首页 云原生大数据计算服务 MaxCompute 操作指南 Data+AI和数据科学 PyODPS 场景实践 PyODPS的空值处理

PyODPS的空值处理

更新时间：
产品详情
我的收藏

本文为您介绍如何使用DataWorks的PyODPS节点处理PyODPS的空值。

前提条件

您需要完成以下操作：

操作步骤

  1. 准备测试数据。
    1. 登录DataWorks控制台。
    2. 创建表并上传数据。操作方法请参见建表并上传数据。
      表pytable2的建表语句如下。 
      CREATE TABLE `pytable2` (
  `id` string,
  `name` string,
  `f1` double,
  `f2` double,
  `f3` double,
  `f4` double
) ;
      示例数据文件pytable2.txt的内容如下。 
      0, name1, 1.0, NaN, 3.0, 4.0
1, name1, 2.0, NaN, NaN, 1.0
2, name1, 3.0, 4.0, 1.0, NaN
3, name1, NaN, 1.0, 2.0, 3.0
4, name1, 1.0, NaN, 3.0, 4.0
5, name1, 1.0, 2.0, 3.0, 4.0
6, name1, NaN, NaN, NaN, NaN

  2. 在左侧导航栏上单击工作空间列表。

  3. 选择操作列中的快速进入 > 数据开发。

  4. 在数据开发页面，右键单击已经创建的业务流程，选择新建节点 > MaxCompute > PyODPS 2。

  5. 在新建节点对话框，输入节点名称，并单击确认。

  6. 在PyODPS 2节点中输入空值处理代码。
    示例代码如下。 
    df2 = DataFrame(o.get_table('pytable2'))
#dropna方法，删除包含空值的行。
print df2.dropna(subset=['f1','f2','f3','f4']).head()

#包含非空值则不删除，可以使用how='all'。
print df2.dropna(how='all', subset=['f1','f2','f3','f4']).head()
print df2.dropna(thresh=3, subset=['f1', 'f2', 'f3', 'f4']).head()

#fillna方法，使用常数或已有的列填充未知值。
print df2.fillna(100, subset=['f1','f2','f3','f4']).head()

#使用一个已有的列填充未知值。
print df2.fillna(df2.f2, subset=['f1','f2','f3','f4']).head()

#向前填充。
print df2.fillna(method='bfill', subset=['f1', 'f2', 'f3', 'f4']).head()

#向后填充。
print df2.fillna(method='ffill', subset=['f1', 'f2', 'f3', 'f4']).head()
  7. 单击运行。运行节点.png
  8. 在运行日志中查看运行结果。运行日志.png
    完整的运行结果如下。 
    Sql compiled:
CREATE TABLE tmp_pyodps_d0c7d8c2_be38_4d48_b0eb_e89bae5bde01 LIFECYCLE 1 AS
SELECT *
FROM WB_BestPractice_dev.`pytable2` t1
WHERE (((IF(t1.`f1` IS NOT NULL, 1, 0) + IF(t1.`f2` IS NOT NULL, 1, 0)) + IF(t1.`f3` IS NOT NULL, 1, 0)) + IF(t1.`f4` IS NOT NULL, 1, 0)) >= 4
Instance ID: 20191010071154980g2hic292

  id   name   f1   f2   f3   f4
0  0  name1  1.0  NaN  3.0  4.0
1  1  name1  2.0  NaN  NaN  1.0
2  2  name1  3.0  4.0  1.0  NaN
3  3  name1  NaN  1.0  2.0  3.0
4  4  name1  1.0  NaN  3.0  4.0
5  5  name1  1.0  2.0  3.0  4.0
6  6  name1  NaN  NaN  NaN  NaN
Sql compiled:
CREATE TABLE tmp_pyodps_49b46768_f589_48f6_be8a_b7139f31f6f2 LIFECYCLE 1 AS
SELECT *
FROM WB_BestPractice_dev.`pytable2` t1
WHERE (((IF(t1.`f1` IS NOT NULL, 1, 0) + IF(t1.`f2` IS NOT NULL, 1, 0)) + IF(t1.`f3` IS NOT NULL, 1, 0)) + IF(t1.`f4` IS NOT NULL, 1, 0)) >= 1

Instance ID: 20191010071159759g0dk9592


  id   name   f1   f2   f3   f4
0  0  name1  1.0  NaN  3.0  4.0
1  1  name1  2.0  NaN  NaN  1.0
2  2  name1  3.0  4.0  1.0  NaN
3  3  name1  NaN  1.0  2.0  3.0
4  4  name1  1.0  NaN  3.0  4.0
5  5  name1  1.0  2.0  3.0  4.0
6  6  name1  NaN  NaN  NaN  NaN
Sql compiled:
CREATE TABLE tmp_pyodps_7f941800_1539_415b_9257_283ebeb893a6 LIFECYCLE 1 AS
SELECT *
FROM WB_BestPractice_dev.`pytable2` t1
WHERE (((IF(t1.`f1` IS NOT NULL, 1, 0) + IF(t1.`f2` IS NOT NULL, 1, 0)) + IF(t1.`f3` IS NOT NULL, 1, 0)) + IF(t1.`f4` IS NOT NULL, 1, 0)) >= 3
Instance ID: 20191010071204544giyswx7

0  0  name1  1.0  NaN  3.0  4.0
1  1  name1  2.0  NaN  NaN  1.0
2  2  name1  3.0  4.0  1.0  NaN
3  3  name1  NaN  1.0  2.0  3.0
4  4  name1  1.0  NaN  3.0  4.0
5  5  name1  1.0  2.0  3.0  4.0
6  6  name1  NaN  NaN  NaN  NaN
Sql compiled:
CREATE TABLE tmp_pyodps_16d6ea6d_5195_4e4c_8346_644a395852f7 LIFECYCLE 1 AS
SELECT t1.`id`, t1.`name`, IF(t1.`f1` IS NULL, 100, t1.`f1`) AS `f1`, IF(t1.`f2` IS NULL, 100, t1.`f2`) AS `f2`, IF(t1.`f3` IS NULL, 100, t1.`f3`) AS `f3`, IF(t1.`f4` IS NULL, 100, t1.`f4`) AS `f4`
FROM WB_BestPractice_dev.`pytable2` t1
Instance ID: 20191010071209190gyl56292

  id   name   f1   f2   f3   f4
0  0  name1  1.0  NaN  3.0  4.0
1  1  name1  2.0  NaN  NaN  1.0
2  2  name1  3.0  4.0  1.0  NaN
3  3  name1  NaN  1.0  2.0  3.0
4  4  name1  1.0  NaN  3.0  4.0
5  5  name1  1.0  2.0  3.0  4.0
6  6  name1  NaN  NaN  NaN  NaN
Sql compiled:
CREATE TABLE tmp_pyodps_40755ebd_2d2a_482e_b360_3f3da0d5422c LIFECYCLE 1 AS
SELECT t1.`id`, t1.`name`, IF(t1.`f1` IS NULL, t1.`f2`, t1.`f1`) AS `f1`, IF(t1.`f2` IS NULL, t1.`f2`, t1.`f2`) AS `f2`, IF(t1.`f3` IS NULL, t1.`f2`, t1.`f3`) AS `f3`, IF(t1.`f4` IS NULL, t1.`f2`, t1.`f4`) AS `f4`
FROM WB_BestPractice_dev.`pytable2` t1

Instance ID: 20191010071213970gbp66792


  id   name   f1   f2   f3   f4
0  0  name1  1.0  NaN  3.0  4.0
1  1  name1  2.0  NaN  NaN  1.0
2  2  name1  3.0  4.0  1.0  NaN
3  3  name1  NaN  1.0  2.0  3.0
4  4  name1  1.0  NaN  3.0  4.0
5  5  name1  1.0  2.0  3.0  4.0
6  6  name1  NaN  NaN  NaN  NaN
Sql compiled:
CREATE TABLE tmp_pyodps_d39fcce1_d8a9_4cc2_8aff_2ed1e9c6bb1b LIFECYCLE 1 AS
SELECT pyodps_udf_1570691538_d9441c59_c666_4a5d_8154_67d8bc8c24ad(t1.`id`, t1.`name`, t1.`f1`, t1.`f2`, t1.`f3`, t1.`f4`) AS (`id`, `name`, `f1`, `f2`, `f3`, `f4`)
FROM WB_BestPractice_dev.`pytable2` t1
Instance ID: 20191010071219627goqv9292


  id   name   f1   f2   f3   f4
0  0  name1  1.0  3.0  3.0  4.0
1  1  name1  2.0  1.0  1.0  1.0
2  2  name1  3.0  4.0  1.0  NaN
3  3  name1  1.0  1.0  2.0  3.0
4  4  name1  1.0  3.0  3.0  4.0
5  5  name1  1.0  2.0  3.0  4.0
6  6  name1  NaN  NaN  NaN  NaN
Sql compiled:
CREATE TABLE tmp_pyodps_3f190cf0_f9fb_4e06_a942_ab31c0241cd3 LIFECYCLE 1 AS
SELECT pyodps_udf_1570691566_0330848b_82d3_411c_88e1_cbbcc6adb9c1(t1.`id`, t1.`name`, t1.`f1`, t1.`f2`, t1.`f3`, t1.`f4`) AS (`id`, `name`, `f1`, `f2`, `f3`, `f4`)
FROM WB_BestPractice_dev.`pytable2` t1
Instance ID: 20191010071247729gt776792

  id   name   f1   f2   f3   f4
0  0  name1  1.0  1.0  3.0  4.0
1  1  name1  2.0  2.0  2.0  1.0
2  2  name1  3.0  4.0  1.0  1.0
3  3  name1  NaN  1.0  2.0  3.0
4  4  name1  1.0  1.0  3.0  4.0
5  5  name1  1.0  2.0  3.0  4.0
6  6  name1  NaN  NaN  NaN  NaN
上一篇：PyODPS的采样下一篇：PyODPS常见问题
文档推荐