# Best practices for complex structured input

**URL:** <https://numba.discourse.group/t/best-practices-for-complex-structured-input/1406>\
**Category:** Numba\
**Created:** [June 23, 2022, 5:30pm UTC](https://numba.discourse.group/t/best-practices-for-complex-structured-input/1406 "2022-06-23T17:30:25Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![jdts](https://yyz2.discourse-cdn.com/free1/user_avatar/numba.discourse.group/jdts/32/635_2.png) [@jdts](https://numba.discourse.group/u/jdts)\
**Post date:** [June 23, 2022, 5:30pm UTC](https://numba.discourse.group/t/best-practices-for-complex-structured-input/1406/1 "2022-06-23T17:30:25Z")

</div>

I’m embarking on an optimization of a long-running `scipy.optimize` code, and hoping to use numba to help. The optimizers allow passing additional arguments to the model function. The model depends on quite a few externally configured scalars and arrays, and I am hoping to package up those arrays into a structured object, like a dict or namedtuple (I see `dataclass` is not yet supported), to pass into my njit’d model function.

As as crude analogy, an argument to a numba-compiled function might look like:

```python
param_map = {"arr1": some_array1, 
             "arr2": {"upper": some_array2, "lower": some_array3},
             "arr3": {"val": 1.0, "other": some_array4}}

```

i.e. a nested list of scalars and arrays, where the `some_array`’s can be 1D numpy arrays, or arrays of numpy structured scalars.

Note that `param_map` is _not_ constant, i.e. it is initialized outside of numba before being passed into the njit’d model function via scipy, and some of its individual values can be updated during the run.

Is there a recommended best practice for this situation, where you have dozens of individual scalars and numpy (structured) arrays you need to pass in as an argument to a numba-njit’d function?