solution #726
solution #726
Conversation
app/main.py
Outdated
|
|
||
| @wraps(func) | ||
| def wrapper(*args) -> Any: | ||
| nonlocal cached, arg_cache |
There was a problem hiding this comment.
this statement is redundant, as dict is mutable type
There was a problem hiding this comment.
Lists are also mutable, so mind this comment
app/main.py
Outdated
|
|
||
| return result | ||
|
|
||
| else: |
There was a problem hiding this comment.
You can use a single return statement out of the if/else statement instead of two inside it.
app/main.py
Outdated
|
|
||
|
|
||
| def cache(func: Callable) -> Callable: | ||
| # Write your code here | ||
| pass | ||
| cached = [] |
There was a problem hiding this comment.
your approach is incorrect: you should store arguments and the result of the function with these arguments and return this result without calling function when arguments are the same.
hint: use dictionary as storage
There was a problem hiding this comment.
I want to use list becouse it is more comfortable to operate list as a row in DataFrame in case if it will be reciered to store as structured info. And both of the list have same input index for ich time use. I saw that a lot of people use dic in this task, but my way also give same output and additionaly more comfortable for Data Scierntist job (in case if it will be required to analize cash)
There was a problem hiding this comment.
You should use a dict here. Under the hood dicts in python use hash tables. You will go through this in Dict Advanced topic but shortly: it works faster than list. You will retrieve values with much less efforts
Using a dictionary provides a more concise and clear representation of the data
It's sutable if you will have a dict instead of two lists for Data Analysis. Data scientists often work with dictionaries or other data types, different from the python lists (numpy arrays for example)
There was a problem hiding this comment.
I used this decorator when train a machine learning model. During the experiment I needed to go through different variants of hyperparameters and correlate these values with the values already obtained. In case of failure to find the optimal parameters, I could use the saved cache values to return to the place where I had started on a non-straight path. In this case, the order of the experiments was also important to me (which could be taken into account in the cache only when using a list)
The task conditions do not say what exactly the caching will be used for, if for the log, then you are right and the speed of information retrieval matters and you should use a dictionary. But if the task is similar to the one I performed with this decorator, then the order of caching is more important than the fact that the result will be 0.0001 seconds slower.
So, since the priority of performance or usability was not explicitly stated, why is my option so unpleasant? And as you say "Data scientists often work with numpy arrays» wich is vary easy to get from Python list, but not dict
app/main.py
Outdated
|
|
||
| return result | ||
|
|
||
| else: |
app/main.py
Outdated
|
|
||
|
|
||
| def cache(func: Callable) -> Callable: | ||
| # Write your code here | ||
| pass | ||
| cached = [] |
There was a problem hiding this comment.
You should use a dict here. Under the hood dicts in python use hash tables. You will go through this in Dict Advanced topic but shortly: it works faster than list. You will retrieve values with much less efforts
Using a dictionary provides a more concise and clear representation of the data
It's sutable if you will have a dict instead of two lists for Data Analysis. Data scientists often work with dictionaries or other data types, different from the python lists (numpy arrays for example)
app/main.py
Outdated
|
|
||
| @wraps(func) | ||
| def wrapper(*args) -> Any: | ||
| nonlocal cached, arg_cache |
There was a problem hiding this comment.
Lists are also mutable, so mind this comment
No description provided.