Hi all!
I’m trying to cache a function generating a pandas dataframe (, which includes an object of the module Bio.pairwise2 of the Bio package. I tried to manually define the hash function for the UnhashableTypeError for the unsupported object which results in a new AttributeError: qualname
This is my first time using caching, so I could imagine a simple answer such as “mutable objects are not hashable”. Is this the case here?
The app: https://share.streamlit.io/jacobhanimann/isoaligner/main/Webinterface.py
Thanks for the help!
Cheers, Jacob
Hi @JacobHanimann,
Thank you for sharing with the Streamlit community!
As the error message notes, you can supply your own function to hash objects of type Bio.pairwise2.alignment_function. The syntax for this looks like the following:
@st.cache(hash_funcs={FooType: hash_foo_type})def my_cached_func(a, b):…
where FooType is the type Streamlit was unable to hash (in this case Bio.pairwise2.alignment_function) and hash_foo_type is a function that properly hashes FooType objects. You can also specify to Streamlit that objects of type Bio.pairwise2.alignment_function should not be hashed via the following:
@st.cache(hash_funcs={Bio.pairwise2.alignment_function: lambda _: None})def my_cached_func(a, b):…
We also provide some helpful caching debugging tips here. Please let me know if you have any questions.
Best,
Caroline
1 Like
