updating multiple attributes python

Question

I have a class Test that has attributes Test.att1, Test.att2, ... , Test.att10. I have a method Test.update() that computes ten new values vals = [v1, v2, ... , v10]. I'd like to update Test's attributes in a single for loop.

I envision something like looping through the list mylist = [self.att1, self.att2, ... , self.att10] within the Test class and setting values per vals but this didn't seem to work. What can I do instead / what's the best way to update multiple attributes without using a self.att1, self.att2, self.att3 = v1, v2, v3 structure?

EDIT: I'm specifically asking because I am doing something like:

a1, a2, a3, a4, a5 = self.update()
self.a1 = pd.concat([self.a1, a1]).drop_duplicates().reset_index(drop=True)
self.a2 = pd.concat([self.a2, a2]).drop_duplicates().reset_index(drop=True)
self.a3 = pd.concat([self.a3, a3]).drop_duplicates().reset_index(drop=True)
self.a4 = pd.concat([self.a4, a4]).drop_duplicates().reset_index(drop=True)
self.a5 = pd.concat([self.a5, a5]).drop_duplicates().reset_index(drop=True)

and am currently using One backwards way I ended up doing this was:

def update(self):
    dfs = [self.att1, self.att2, ..., self.att10]
    dfs = [pd.concat([dfs[i], vals[i]]).drop_duplicates().reset_index(drop=True) \
                for i in range(len(dfs))]
    [self.att1, self.att2, ..., self.att10] = dfs

but at seeking a better way

Why don't you write a method which will take the list and update the attributes? it will be easier for you in that way. — Max
– Max, Commented Sep 7, 2017 at 4:31
They all share a similar structure along the lines of pd.concat([self.att1,df]).drop_duplicates().reset_index(drop=True) so I was wondering how to only write once and avoid writing 10 times... — jjjjjj
– jjjjjj, Commented Sep 7, 2017 at 4:34

Nelson · Accepted Answer · 2017-09-07 22:22:55Z

1

You could use setattr in combinatin with hasattr to update and add attributes to the target object.

attrs = [a for a in dir(Test) if a.startswith('attr') and 'no_update' not in a]

### To update existing attrs ###
new_vals = [...]
for a, v in zip(attrs, new_vals):
    if hasattr(Test, a):
        setattr(Test, a, v)

### To add new attrs ###
new_vals = [...]
start = int(attrs[-1][4:]) + 1 # the number the new attrs should start at
new_attrs = ['attr%d' % n for n in range(start, start+len(new_vals))] # if start==4 and len(new_vals)==3: return ['attr5', 'attr6', 'attr7']
for a, v in zip(new_attrs, new_vals):
    setattr(Test, a, v)

These can also be combined to allow updating existing attrs and setting new ones if need be.

edited Sep 7, 2017 at 22:22

answered Sep 7, 2017 at 4:42

Nelson

9521 gold badge9 silver badges25 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

jjjjjj Over a year ago

Thanks! What if I want to append v to attribute a as per my update? Is this still a clean way of doing it?

Nelson Over a year ago

By that do you mean concatenating the new and old attr value? If so, you can get the attr just like you can set and check if it has it like so: setattr(Test, a, getattr(Test, a) + v)

jjjjjj Over a year ago

Basically these attributes are pandas DataFrames, and I want to add new lines to an existing 10 or so DataFrames which I have as attributes of Test, but I also have other attributes att_no_update1, ... att_no_update10 that I don't want to update

Nelson Over a year ago

@jjjjjj I have no experience with pandas and have no idea what a DataFrame is. Regardless, I think I understand what you want and have updated my answer accordingly

Collectives™ on Stack Overflow

updating multiple attributes python

1 Answer 1

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related