Aidemy https:// aidemy.net 2020/9/21
Nice to meet you, this is Ngahope! Although I am a liberal arts student, I started attending the AI learning school "Aidemy" this week because I was interested in the possibilities of AI. There, I am learning various knowledge about AI. I wanted to share this knowledge with you, so I decided to put it together on Qiita.
Please read it!
What to learn this time ・ One line notation ・ Split display of list ・ Processing performed on the dictionary
lambda（1） -A function defined by def and returned by return can be expressed in one line by using lambda.
__ · lambda Argument: Return value __
# x*Define a function a that returns 4
a=lambda x: x*4
print(a(4))
# 16
-Just separate the arguments with ","
a=lambda x,y: x+y
print(a(3,6))
# 9
・ Write if else side by side __ · lambda Argument: Process (True) if condition else Other process __
a=lambda x: x*4 if x>4 else x=4
print(a(6))
# 24
-Use the split function when you want to split by a certain symbol (one type). __ · String.split ("symbol") __
text="my name is ngayope"
text.split(" ") # " (space)"Separated by
# ['my','name','is','ngayope']
-Use the re.split function when you want to split by multiple symbols (re must be imported to use) __ ・ re.split ("[symbol]", character string) __
import re
text="Bulbasaur,Charmander.Squirtle"
re.split("[,.]",text)
# ['Bulbasaur','Charmander','Squirtle']
-When using a function for each element of the list, use the map function. Such a function is called an "iterator". __ ・ list (map (function, list)) __ * If you do not enclose it in list (), the function application result will not be reflected!
import re
time_list =["2006/11/26_2:40","2009/1/16_23:35","2014/5/4_14:26","2017/8/9_7:5","2017/4/1_22:15"]
# time_Function to extract "time" from list
hour_pick = lambda x: int(re.split("[/_:]",x)[3])
#↑ First re.Split the string with split. Then int this string()Convert to a number with. Finally[3]Display the function to extract the "hour" part with lambda in one line.
#time_Apply the function to list and return the result.
list(map(hour_pick,time_list))
# [2,23,14,7,22]
-It can be extracted by using the filter function in the same way as map. __ · list (filter (authenticity judgment function, list)) __
#Above time_Function that sets True only for the list whose "month" is after July
judge=lambda a: int(re.split("[/_:]",a)[1] >6
list(filter(judge,time_list))
# ["2006/11/26_2:40","2017/8/9_7:5"]
-Use the sorted function instead of the sort function. __ · sorted (list, key = reference function, reverse = True (descending order)) __
list1=[[2,3][4,1][5,5][9,0][0,7][1,6]]
#Sort based on the second element (descending order)
sorted(list1,key=lambda x: x[1],reverse=False)
# [[0,7][1,6][5,5][2,3][4,1][9,0]]
-Even if you do the following, you can apply the function to all the elements in the same way as the map function. __ · [Function for Variables in List] __
cm=[100,50,500,3,380]
#[◯m,◯cm]Function to calculate like
m_cm=lambda x:[x//100,x%100]
print([m_cm(x) for x in cm])
#[[1,0],[0,50],[5,0],[0,3],[3,80]]
-Even if you do the following, you can extract only the elements that satisfy the conditions like the filter function. __ · [Element for Variable in List if Authenticity Function] __
#Extract only "things over 1m" in the upper cm
[x for x in cm if x>=100]
# [100,500,380]
-Even if it is a separate list, one loop processing can be performed using the zip function. __ · for variables 1,2 in zip (Listing 1,2): __ processing __ ・ [Processing for variables 1, 2 in zip (List 1, 2)] __
a = [1,-2,3,-4,5]
b = [9,8,-7,-6,-5]
#For each element of the list, x*4+y*Do 2
[x*4+y*2 for x,y in zip(a, b)]
# [22,8,-2,-28,10]
・ When looping one of the two lists and looping the other -Set a for statement in each of the two lists and describe the processing for the two variables. -Use the following method for in-list processing. ** [[Process] for Variable 1 in List 1 for Variable 2 in List 2] **
a=[1,2]
b=[5,6]
print([[x+y]for x in a for y in b])
# [6,7,7,8]
-Normally, it is necessary to initialize the contents of the dictionary one by one, but it can be easily added by using the defaultdict class instead of the dictionary. ** ・ defaultdict (element type (int, dir, etc.)) **
from collections import defaultdict
d=defaultdict(int)
lst=['Bulbasaur','Charmander','Squirtle','Charmander']
#Extract each element of the list as a key and increase the number of keys that appear by +1
for key in lst:
__d[key] += 1
print(d)
# defaultdict(<class='int'>,{'Bulbasaur':1,'Charmander':2,'Squirtle':1})
-You can easily add elements by using defaultdict. ** ・ defaultdict dictionary [key] .append (element) **
#list type dictionary
a=[('Charmander',5),('Lizard',16),('Lizardン',36),('Lizardン',100)]
d=defaultdict(list)
#index of a[0]To x, index[1]Is assigned to y and taken out
for x,y in a:
__d[x].append(y)
print(d)
#defaultdict(<class='list'>,{'Charmander':[5],'Lizard':[16],'Lizardン':[36,100]})
-By using the Counter class, it is easier to count than the default dict. ** ・ Counter (counting data) ** -By setting ".most_common (number of elements)", sort in descending order and output the specified number of elements.
from collections import Counter
lst=['Bulbasaur','Charmander','Squirtle','Charmander']
print(Counter(lst).most_common(1))
# Counter({'Charmander':2})
-By using lambda, the function is multiplied by one line. ・ Split and re. You can split the list with split. -If you write a for statement and a function in the list [], you can easily apply the function to the elements of the list. -By using defaultdict and Counter, you can easily count the elements of the dictionary.
This time is over. Thank you for reading this far.
Recommended Posts