Handcalcs with pandas and numpy

Currently, handcalcs is not currently compatible with pandas and numpy. While it is relatively easy to render a numpy array, matrix, or pandas dataframe as a latex table (even one compatible with MathJax and Katex), I am not yet clear on the best way of representing these objects as part of the handcalcs calculation procedure.

Handcalcs uses a parameter = symbolic vars = numeric substitution = result style of representation. How would this work with arrays? Arrays much larger than 5x5 are particularly a problem (e.g. do we just use three dots in the middle rows/cols? Do we just use .head()? For column operations, do we show columns as vertical arrays separated by operators? Could we just use column names or array indexes? How do we represent operations such as division? How much information is enough?) Additionally,handcalcs cannot handle working with slices because that would require evaluating a slice prior to rendering the results of the slice and handcalcs does not do any evaluations.

New ideas, in the form of Latex code, are needed!