How Spark SQL columns convert data types

Spark Big Data Development Practice 5: Spark RDD-KV Type Value Conversion.

To modify the type, use the SQL statement

alter table name alter column field name varchar not null

Example: Create a table as shown in the following figure.

Shown: Change the name column to varchar with a length of 20, as shown in the following figure:

After the change, the result is:

alter table table name alter column column name new data type [(length)] null or not null

Example: alter table teacher alter column office char(20) not null default"office"

If the length is sufficient, it can be used.

alter table name.

alter column field name: nvarchar(255), if the length is not enough, you need to add a column, and then convert the source data column to the length and put it in.

reatetable

mytest

fidint?What about the pavilion? Knock Chun Squid and stare at Damin? Yun, Fnamenvarchar (20).

null) to create a table named . If you want to change the cavity at this point.

fname column: mytest, which contains two fields? slug, then you can use the following sentence Kaizhi:

altertable

mytest

Reed? Monkey? Ugly? Imperial Prying? Neon.

You can use alter table plus alter column to change the specified column data type, for example:

create table

mytest (fid int,fname nvarchar(20) null)

Create a table, named mytest, with two fields. If you want to change the fname column at this point, you can use the following statement:

alter table mytest

alter column fname nvarchar(30) null

It should be noted here that if the specified new data type is smaller in length than the original data type, then there will be data loss, so be aware that in addition, not all data types can be changed, the following are the precautions mentioned in the SQL Online series:

The column you want to change can't be:

Columns with data type text, image, ntext, or timestamp.

The rowguidcol column of the table.

Calculated columns or columns used to calculate columns.

The column is being copied. Columns used in indexes, unless the column data type is varchar, nvarchar, or varbinary, the data type has not changed, and the new column size is equal to or greater than the old column size.

A column used in statistics created by the create statistics statement. First, delete the statistics with the drop statistics statement. Statistics automatically generated by the query optimizer are automatically removed by the alter column.

Columns used in the Primary Key or [Foreign Key] References constraints.

Columns used in check or unique constraints, unless the length of variable-length columns used in check or unique constraints allows change.

Columns that have default values associated with them, unless you are allowed to change the length, precision, or number of decimal places of the column without changing the data type.

alter

table name: alter

column

The column name is the new data type [(length)].

null or not

null example: alter

Table teacher alter

column

Office char (20).

notnull

default

office"Put the written sentence in a place where other SQL statements are executed, such as execute (SQL), execute it once, and then change it back.

alter table table name alter column column name new data type [(length)] null or not null

Example: alter table teacher alter column office char(20) not null default"office"

Put the written sentence in a place where other SQL statements are executed, such as execute (SQL), execute it once, and then change it back.

It doesn't seem necessary. What is the same thing between the nchar type and the nvarchar type? For conversion with other multiple characters, such as Chinese, phonetic transcription, etc., 2 bytes are occupied for each English (ASCII) character, 2 bytes for one Chinese character, and 2 bytes for all characters.