更改 pandas csv 列的值样式 python

Question

I have a dataset like below:

my problem is in anotation column, I want to change the style of lists that is showing above into some thing like this:

• ['flight-sth sth sth sth]

I mean that the anotation column value's has multiple style, I just want to change that style into mine:)
example:

['flight_search.price_range'] ==> ['flight-search price range']  
['flight_search.stops'] ==> ['flight-search stop']  
['flight_search.date.depart_origin'] ==> ['flight-search date depart origin']  

and after doing this conversion, replacing it exactly into old anotation column:)

---

the sample of anotation:

anotation
['flight_search.destination1']  
['flight_search.origin']  
['flight_search.destination1']  
['flight_search.type']  
['flight_search.type']  
['flight_search.airline']  
['flight_search.stops']  
['flight_search.stops']  
['flight_search.price_range']  
['flight_search.price_range']  
['flight1_detail.from.time']  
['flight_search.date.depart_origin']  

Answer 1

annotation = [['flight_search.destination1'],  
['flight_search.origin'],
['flight_search.destination1']  ,
['flight_search.type']  ,
['flight_search.type']  ,
['flight_search.airline'],  
['flight_search.stops']  ,
['flight_search.stops']  ,
['flight_search.price_range']  ,
['flight_search.price_range']  ,
['flight1_detail.from.time']  ,
['flight_search.date.depart_origin']]  

empty = []
for i in annotation:
    empty.append([i[0].replace("_","-").replace("."," ")])

Output

[['flight-search destination1'],
 ['flight-search origin'],
 ['flight-search destination1'],
 ['flight-search type'],
 ['flight-search type'],
 ['flight-search airline'],
 ['flight-search stops'],
 ['flight-search stops'],
 ['flight-search price-range'],
 ['flight-search price-range'],
 ['flight1-detail from time'],
 ['flight-search date depart-origin']]

DataFrames

# for dataframe

df["annotation"].apply(lambda x: [x[0].replace("_","-").replace("."," ")])

I Believe this should do the trick, if no typos are in it

Answer 2

Python String replace() Method could be an option. But i see you wanted the first underscore to be a - and the second one being a space. I think that problem can be solved if you go deep into regular expressions in python. to keep it simple i've made this so far:

mystring = 'flight_search.price_range'
mystring = mystring.replace("_", "-")
mystring = mystring.replace(".", " ")

see https://www.w3schools.com/python/ref_string_replace.asp

edited code:

mystring = 'flight_search.price_range'
mystring = mystring.replace("_", "-",1)
mystring = mystring.replace(".", " ")
mystring = mystring.replace("_", " ")
print(mystring)

result of edited code: flight-search price range

Answer 3

What you want to think about is what changes do you need to make to the strings in the annotation column. With the df.replace() function you can apply simple changes to all the columns.

If you need some more control, however, you would want to use the df.apply() function. With this function you can specify exactly what you want to do with each string in the column using a custom function.

For example you could take this approach to start off with, you can change the custom function to get your desired results:

import pandas as pd

annotation = ['flight_search.destination1',  
'flight_search.origin',
'flight_search.destination1',
'flight_search.type' ,
'flight_search.type'  ,
'flight_search.airline',  
'flight_search.stops'  ,
'flight_search.stops'  ,
'flight_search.price_range' ,
'flight_search.price_range' ,
'flight1_detail.from.time' ,
'flight_search.date.depart_origin']

df = pd.DataFrame({"annotation":annotation})

def custom_func(string):
    # replace the initial word
    string = string.replace("flight_", "flight-")
    string = string.replace("flight1_", "flight1-") # is this a typo?
    
    # replace the other punctuataion marks with a space
    for punctuation in ['_', '.']:
        string = string.replace(punctuation, " ")
    
    # retun the formatted string
    return string

# apply the custom function to the annotation column
df["annotation"] = df["annotation"].apply(custom_func)