Defaultdict example does not work (for me)

[mgandaman]

Hello! First of all, thank you for this excellent tutorial! I have been learning Python like stumbling around in a dark room full of dumbbells and this tutorial has helped me immensely so far!

I am on "Data Structures (Part III)" and encountered an issue when I tried running the examples for defaultdict. I cannot get any of them to run. When I run the default code I get an error as follows:

This line (Line 4 in my jupyter notebook cell):
example_default_dict = defaultdict(list) # will map any missing key to list()

Error message:
TypeError: first argument must be callable or None

I am running the code in a Jupyter notebook running Python version 3.7.4. I am unsure what I am doing wrong - is it an issue with my version of Python? The examples I found online indicate that I need to define a variable as a lambda and then use that variable where you have "list" in defaultdict. I am new to this so any help you can provide is much appreciated.

Thank you!

[rsokl]

Hi @mgandaman thank you for taking the time to reach out and troubleshoot this! I am very glad that PLYMI has been helpful so far 😄

There is certainly nothing wrong with your version of Python. I am wondering if perhaps there was a typo in your code when you ran it. Could you double check to see that you didn't accidentally write:

 from collections import defaultdict
 example_default_dict = defaultdict(list())  # note that this is `list()` and not `list`

This indeed will raise the TypeError message that you found.

The difference between passing list and list() is:

• list is a function (a.k.a a "callable") that defaultdict can use to make lists as-needed for missing key-value pairs.
• list() (that is calling the list function with no inputs) simply returns an empty list. An empty list is not a function (i.e. not a callable) and thus defaultdict can't use it to create values. Thus it raises the aforementioned TypeError.

And note that, because list is already a callable, there is no need to use lambda anywhere here. A lambda expression is just a convenient way to define one-liner Python functions.

Please let me know if this addresses your issue!

And FYI, in order to include formatted code in a GitHub post, use the formatting:

```python
# your code here
x = 2 + 3
```

[mgandaman]

Thank you for the help!
Here is the code as I entered it in the notebook:

# demonstrate the behavior of the `defaultdict`
from collections import defaultdict

example_default_dict = defaultdict(list)  # will map any missing key to `list()`
example_default_dict  # an empty default dictionary
defaultdict(list, {})

# "apple" is not a key, so the default mapping "apple" -> list() is created
# and this value is returned
example_default_dict["apple"]

# this mapping now exists in the dictionary
example_default_dict

I've also included a screenshot of the code and the error messages as they appear.
P.S. thanks for the pointers, I'm new to GitHub!

[rsokl]

This is strange... Is there any chance that you accidentally overwrote the definition of list in your notebook? Could you try restarting your notebook's kernel and re-running that single cell?

I copied your code as-is and ran it in my notebook without any issue.

[mgandaman]

Restarting the kernel fixed it! I searched through the notebook and didn't see list defined anywhere, but it seems that was the problem. Thank you!

[rsokl]

Great! Yeah that was a really subtle problem. This is a downside to using Jupyter notebooks: you can write and run code, and then delete it - leaving no trace!

Please share any feedback you have as you continue through PLYMI - I am always looking to improve it!