I challenged you π...
You delivered πͺ!
I asked you to implement the sum function in Python π.
Now I'll go over some alternatives and tell you what I like β and dislike β about them.
I'll also tell you which one I think is the best, most Pythonic β¨ one.
π
You delivered πͺ!
I asked you to implement the sum function in Python π.
Now I'll go over some alternatives and tell you what I like β and dislike β about them.
I'll also tell you which one I think is the best, most Pythonic β¨ one.
π
I didn't get many submissions, so this is going to be a short thread.
The first approach you might feel tempted to follow is by using a `for` loop.
Without too much effort, you can write a pretty clean solution π
The first approach you might feel tempted to follow is by using a `for` loop.
Without too much effort, you can write a pretty clean solution π
What's good about this solution is that, with no extra effort, it supports the βoddβ cases that the built-in `sum` already supports.
For example, I've shown that the built-in `sum` can be used to flatten lists.
So can this implementation:
For example, I've shown that the built-in `sum` can be used to flatten lists.
So can this implementation:
But there's something else in here.
There's an idea that I'd like to explore.
And it's an idea I wanted you to figure out by yourselves:
There's an idea that I'd like to explore.
And it's an idea I wanted you to figure out by yourselves:
https://twitter.com/mathsppblog/status/1454095879523217417
`sum`, alongside `all`, `any`, `min`, `max`, `math.prod`, and others is from a very specific family of algorithms!
`sum` is what is called a reduction!
Given a sequence and an operation, a reduction uses that operation to reduce the sequence into a single value.
`sum` is what is called a reduction!
Given a sequence and an operation, a reduction uses that operation to reduce the sequence into a single value.
In the case of `sum`, the operation being used is the addition.
We take a sequence, and reduce it to a single value by summing all the values in the sequence.
In the case of `math.prod`, the operation is multiplication.
For `all`, the operation is `and`.
Etc...
We take a sequence, and reduce it to a single value by summing all the values in the sequence.
In the case of `math.prod`, the operation is multiplication.
For `all`, the operation is `and`.
Etc...
So, once we recognise that `sum` is a reduction, we can actually implement it as such.
That's why many of you suggested an implementation similar to this one π
That's why many of you suggested an implementation similar to this one π
By using the `operator` module we don't even need to define a `lambda` that wraps `+`, and our `sum` implementation becomes much more direct and to the point.
Now, many of you will argue that the `for` loop is better because it is simpler...
What do you think? ππ¬
Now, many of you will argue that the `for` loop is better because it is simpler...
What do you think? ππ¬
I prefer to use `reduce` because it tells you, right from the start, the type of algorithm that you used to implement the function.
I like `reduce` so much, I even wrote a whole article explaining how it works, and why it is useful to know about it:
mathspp.com/blog/pydonts/tβ¦
I like `reduce` so much, I even wrote a whole article explaining how it works, and why it is useful to know about it:
mathspp.com/blog/pydonts/tβ¦
β’ β’ β’
Missing some Tweet in this thread? You can try to
force a refresh
γ
