current position:Home>Glom module of Python data analysis module (1)

Glom module of Python data analysis module (1)

2022-01-30 13:48:33 cxapython

Little knowledge , Great challenge ! This article is participating in “ A programmer must have a little knowledge ” Creative activities .

This article has participated in  「 Digging force Star Program 」 , Win a creative gift bag , Challenge creation incentive fund

Under the simple said glom The module is mainly used for processing structured data , Simple installation pip install  glom that will do , Next is glom Explain the method parameters with examples .


With the same name as the module glom Method of use :

.glom(target, spec, **kwargs)
target Parameters are structured data , It's usually json Nested Type .
spec Parameters are defined templates
**kwargs Contains parameters of key value type :default,skip_exc,scope

Let's take a simple example 、

target = {'a': {'b':{'c':{'d':{'e':[1,2,3,4,5,6,7]}}}}}
spec = 'a.b.c.d.e'
output = glom(target, spec)
# Output [1, 2, 3, 4, 5, 6, 7]
 Copy code 

Let's try more games :

target = {'a': {'b':{'c':{'d':{'e':[1,2,3,4,5,6,7]}}}}}
spec = {'aa':'a.b.c.d.e'}
output = glom(target, spec)
# Output {'aa': [1, 2, 3, 4, 5, 6, 7]}

target2 = {'a': {'b':{'c':[{'d':[1,2,3,4]},{'d':[4,5,6,7]}]}}}
spec2 = {'test':('a.b.c',['d'])}
output2=glom(target2, spec2)
# Output {'test': [[1, 2, 3, 4], [4, 5, 6, 7]]}
 Copy code 

Another official example , Make comprehensive use of :

target = {'system': {'planets': [{'name': 'earth', 'moons': 1},
                                 {'name': 'jupiter', 'moons': 69}]}}

spec = {'names': ('system.planets', ['name']),
            'moons': ('system.planets', ['moons'])}
 Output :{'moons': [1, 69], 'names': ['earth', 'jupiter']}
 Copy code 

As can be seen from the above examples ,spec The basic usage is '' Write the key names in turn . Key name . Key name . Key name ,

Another thing to note , If there is a dictionary in the list to take its value, the key name must be the same, just above 'd' key .

It can be seen that glom and json It saves a lot of []. Easy to use , in addition spec The expression also supports the use of expressions, such as lambda.
Let's take a look at an example with complete parameters

target = {'a': [0, 1, 2]} # Given a structured data 
spec = {'a': ('a', [lambda x:x/x])}# obtain a The value of each element divided by its own value .
output = glom(target, spec,default='666',skip_exc=ZeroDivisionError)
# Output 666
 Copy code 

The above results mainly demonstrate spec The advanced usage of and the usage of the third parameter . The above calculation 0/0 It's a mistake , Wrong type ZeroDivisionError, So we let ZeroDivisionError Error skipped , Simultaneous coordination default The default value when given an error .
But what if the following happens ?

t={'a': {'b.c':{'c':{'d':{'e':[1,2,3,4,5,6,7]}}}}}
s= {'aa':'a.b.c.c'}
 Copy code 

Now let's see Path Method


Path objects specify explicit paths when the default 'a.b.c'-style general access syntax won’t work or isn’t desirable. Use this to wrap ints, datetimes, and other valid keys, as well as strings with dots that shouldn’t be expanded, Let's look at an example .

from glom import glom,Path
target2 = {'a': {'b':{'c':[{'d':[1,2,3,4]},{'d':[4,5,6,7]}],'w':{'3.c':'e'}}}}
# Output e
 Copy code 

Then let's go back and look at the use of spec Errors in the way

t={'a': {'b.c':{'c':{'d':{'e':[1,2,3,4,5,6,7]}}}}}
# Output {'aa': {'d': {'e': [1, 2, 3, 4, 5, 6, 7]}}}
 Copy code 

Be careful :

1.path The hierarchical relationship of the path must be next to . For example, I directly Path('a','c') It's going to be a mistake .
2.path You can't take the keys in the list like this. You need to cooperate glom Handle .
Path('system','planets','name') perhaps Path('system','planets',['name']) Neither. .

target = {'system': {'planets': [{'name': 'earth', 'moons': 1},
                                 {'name': 'jupiter', 'moons': 69}]}}

 follow-up ...
 Copy code 

copyright notice
author[cxapython],Please bring the original link to reprint, thank you.

Random recommended