用 Python 定義 Schema 并生成 Parquet 文件詳情

更新時間：2021年09月25日 10:19:40 作者：Yanbin Blog

本文將演示兩個例子，一個是沒有層級的兩個字段，另一個是含于嵌套級別的字段，將要使用到的 Python 模塊有 pandas 和 pyarrow，感興趣是我小伙伴請和小編一起學(xué)習(xí)下面文章內(nèi)容吧

一、簡單字段定義

1、定義 Schema 并生成 Parquet 文件

import pandas as pd
import pyarrow as pa
import pyarrow.parquet as pq

# 定義 Schema
schema = pa.schema([
    ('id', pa.int32()),
    ('email', pa.string())
])

# 準(zhǔn)備數(shù)據(jù)
ids = pa.array([1, 2], type = pa.int32())
emails = pa.array(['first@example.com', 'second@example.com'], pa.string())

# 生成 Parquet 數(shù)據(jù)
batch = pa.RecordBatch.from_arrays(
    [ids, emails],
    schema = schema
)
table = pa.Table.from_batches([batch])

# 寫 Parquet 文件 plain.parquet
pq.write_table(table, 'plain.parquet')
import pandas as pd

import pyarrow as pa

import pyarrow . parquet as pq

# 定義 Schema

schema = pa . schema ( [

     ( 'id' , pa . int32 ( ) ) ,

     ( 'email' , pa . string ( ) )

] )

# 準(zhǔn)備數(shù)據(jù)

ids = pa . array ( [ 1 , 2 ] , type = pa . int32 ( ) )

emails = pa . array ( [ 'first@example.com' , 'second@example.com' ] , pa . string ( ) )

# 生成 Parquet 數(shù)據(jù)

batch = pa . RecordBatch . from_arrays (

     [ ids , emails ] ,

     schema = schema

)

table = pa . Table . from_batches ( [ batch ] )

# 寫 Parquet 文件 plain.parquet

pq . write_table ( table , 'plain.parquet' )

2、驗證 Parquet 數(shù)據(jù)文件

我們可以用工具 parquet-tools 來查看 plain.parquet 文件的數(shù)據(jù)和 Schema

 $ parquet-tools schema plain.parquet  message schema {      optional int32 id;      optional binary email (STRING);  }  $ parquet-tools cat --json plain.parquet  {"id":1,"email":"first@example.com"}  {"id":2,"email":"second@example.com"}

沒問題，與我們期望的一致。也可以用 pyarrow 代碼來獲取其中的 Schema 和數(shù)據(jù)

schema = pq.read_schema('plain.parquet')
print(schema)

df = pd.read_parquet('plain.parquet')
print(df.to_json())
schema = pq . read_schema ( 'plain.parquet' )

print ( schema )

df = pd . read_parquet ( 'plain.parquet' )

print ( df . to_json ( ) )

輸出為：

schema = pq.read_schema('plain.parquet')
print(schema)

df = pd.read_parquet('plain.parquet')
print(df.to_json())
schema = pq . read_schema ( 'plain.parquet' )

print ( schema )

df = pd . read_parquet ( 'plain.parquet' )

print ( df . to_json ( ) )

二、含嵌套字段定義

下面的 Schema 定義加入一個嵌套對象，在 address 下分 email_address 和 post_address，Schema 定義及生成 Parquet 文件的代碼如下

import pandas as pd
import pyarrow as pa
import pyarrow.parquet as pq

# 內(nèi)部字段
address_fields = [
    ('email_address', pa.string()),
    ('post_address', pa.string()),
]

# 定義 Parquet Schema，address 嵌套了 address_fields
schema = pa.schema(j)

# 準(zhǔn)備數(shù)據(jù)
ids = pa.array([1, 2], type = pa.int32())
addresses = pa.array(
    [('first@example.com', 'city1'), ('second@example.com', 'city2')],
    pa.struct(address_fields)
)

# 生成 Parquet 數(shù)據(jù)
batch = pa.RecordBatch.from_arrays(
    [ids, addresses],
    schema = schema
)
table = pa.Table.from_batches([batch])

# 寫 Parquet 數(shù)據(jù)到文件
pq.write_table(table, 'nested.parquet')
import pandas as pd

import pyarrow as pa

import pyarrow . parquet as pq

# 內(nèi)部字段

address_fields = [

     ( 'email_address' , pa . string ( ) ) ,

     ( 'post_address' , pa . string ( ) ) ,

]

# 定義 Parquet Schema，address 嵌套了 address_fields

schema = pa . schema ( j )

# 準(zhǔn)備數(shù)據(jù)

ids = pa . array ( [ 1 , 2 ] , type = pa . int32 ( ) )

addresses = pa . array (

     [ ( 'first@example.com' , 'city1' ) , ( 'second@example.com' , 'city2' ) ] ,

     pa . struct ( address_fields )

)

# 生成 Parquet 數(shù)據(jù)

batch = pa . RecordBatch . from_arrays (

     [ ids , addresses ] ,

     schema = schema

)

table = pa . Table . from_batches ( [ batch ] )

# 寫 Parquet 數(shù)據(jù)到文件

pq . write_table ( table , 'nested.parquet' )

1、驗證 Parquet 數(shù)據(jù)文件

同樣用 parquet-tools 來查看下 nested.parquet 文件

 $ parquet-tools schema nested.parquet  message schema {      optional int32 id;      optional group address {          optional binary email_address (STRING);          optional binary post_address (STRING);      }  }  $ parquet-tools cat --json nested.parquet  {"id":1,"address":{"email_address":"first@example.com","post_address":"city1"}}  {"id":2,"address":{"email_address":"second@example.com","post_address":"city2"}}

用 parquet-tools 看到的 Schama 并沒有 struct 的字樣，但體現(xiàn)了它 address 與下級屬性的嵌套關(guān)系。

用 pyarrow 代碼來讀取 nested.parquet 文件的 Schema 和數(shù)據(jù)是什么樣子

schema = pq.read_schema("nested.parquet")
print(schema)

df = pd.read_parquet('nested.parquet')
print(df.to_json())
schema = pq . read_schema ( "nested.parquet" )

print ( schema )

df = pd . read_parquet ( 'nested.parquet' )

print ( df . to_json ( ) )

輸出：

id: int32
  -- field metadata --
  PARQUET:field_id: '1'
address: struct&lt;email_address: string, post_address: string&gt;
  child 0, email_address: string
    -- field metadata --
    PARQUET:field_id: '3'
  child 1, post_address: string
    -- field metadata --
    PARQUET:field_id: '4'
  -- field metadata --
  PARQUET:field_id: '2'
{"id":{"0":1,"1":2},"address":{"0":{"email_address":"first@example.com","post_address":"city1"},"1":{"email_address":"second@example.com","post_address":"city2"}}}
id : int32

   -- field metadata --

   PARQUET : field_id : '1'

address : struct & lt ; email_address : string , post_address : string & gt ;

   child 0 , email_address : string

     -- field metadata --

     PARQUET : field_id : '3'

   child 1 , post_address : string

     -- field metadata --

     PARQUET : field_id : '4'

   -- field metadata --

   PARQUET : field_id : '2'

{ "id" : { "0" : 1 , "1" : 2 } , "address" : { "0" : { "email_address" : "first@example.com" , "post_address" : "city1" } , "1" : { "email_address" : "second@example.com" , "post_address" : "city2" } } }

數(shù)據(jù)當(dāng)然是一樣的，有略微不同的是顯示的 Schema 中, address 標(biāo)識為 struct<email_address: string, post_address: string> , 明確的表明它是一個 struct 類型，而不是只展示嵌套層次。

到此這篇關(guān)于用 Python 定義 Schema 并生成 Parquet 文件詳情的文章就介紹到這了,更多相關(guān)用 Python 定義 Schema 并生成 Parquet 文件內(nèi)容請搜索腳本之家以前的文章或繼續(xù)瀏覽下面的相關(guān)文章希望大家以后多多支持腳本之家！

您可能感興趣的文章: