create a MinHash function to copy abundances from one MinHash to another

[ctb]

Over in #1392, I implement a code block that retrieves abundances from a sketch using hashes; in src/sourmash/commands.py, function gather:

            if is_abundance:
                # reinflate abundances
                hashes = set(next_query.minhash.hashes)
                orig_abunds = orig_query_mh.hashes
                abunds = { h: orig_abunds[h] for h in hashes }

                abund_query_mh = orig_query_mh.copy_and_clear()
                # orig_query might have been downsampled...
                abund_query_mh.downsample(scaled=next_query.minhash.scaled)
                abund_query_mh.set_abundances(abunds)
                next_query.minhash = abund_query_mh

Here, next_query.minhash is a flattened MinHash object derived from orig_query_mh, and we are constructing a new MinHash object abund_query_mh that has all of the hashes from next_query.minhash with abundances for those hashes taken from orig_query_mh.

This is potentially a generically useful function that is kind of the converse of flatten(), so let's make it a method on the MinHash class!

TODO:

• create a new method on MinHash objects (see src/sourmash/minhash.py) named something like inflate;
• this method should take one argument in addition to self, a MinHash object with track_abundance=True;
• this method should create a new MinHash object that borrows the abundances from the second argument using only the hashes from self;
• replace the code block in commands.gather() with using this new function (and see if it can apply to multigather;
• write some tests for tests/test_minhash.py;

[keyabarve]

@ctb Is this inflate function supposed to be written in src/sourmash/minhash.py? Should this function be inside any particular class?

[ctb]

yes, in the MinHash class. It probably needs to be defined on FrozenMinHash where it should raise an exception, too.

[ctb]

(order doesn't matter)