io.read_metadata does not always set index dtype as expected

[joverlee521]

Current Behavior

We try to explicitly set the dtype for the index column within io.read_metadata:

augur/augur/io.py

Lines 116 to 121 in 139ba04

	# If we found a valid column to index the DataFrame, specify that column and
	# also tell pandas that the column should be treated like a string instead
	# of having its type inferred. This latter argument allows users to provide
	# numerical ids that don't get converted to numbers by pandas.
	kwargs["index_col"] = index_col
	kwargs["dtype"] = {index_col: "string"}

However, this does not work as expected because of a bug in pandas where the dtype is ignored for index_col.

Additional context

Found this issue when looking into nextstrain/ncov#948

[victorlin]

Seems like this was fixed in pandas 1.4.0, should we bump minimum requirement in setup.py?

[joverlee521]

Hmm, starting with pandas 1.4.0, they only support Python 3.8 and higher. Maybe we should just work around it to avoid bumping the minimum Python dependency for Augur.