Let pygmt.info support pandas.DataFrame or 2D numpy.array inputs

[weiji14]

Description of the desired feature

The pygmt.info function should accept pandas.DataFrame or 2D numpy.array matrices as input. This was mentioned in passing when pygmt.info was originally implemented at #106 (comment). Currently it can only handle filename (str) types:

pygmt/pygmt/modules.py

Lines 89 to 90 in bedb48d

	if not isinstance(fname, str):
	raise GMTInvalidInput("'info' only accepts file names.")

What I'm particularly interested in is getting a nicely rounded xmin/xmax/ymin/ymax 'region' by simply passing in a pandas.DataFrame (Option 3 in the related issue at #147). Something like this:

import pandas as pd

df: pd.DataFrame = pygmt.datasets.load_ocean_ridge_points()
region = pygmt.info(df, spacing="0.1")  # pygmt.info("@ridge.txt", I="0.1")
print(region)
# would output '-R-180/180/-65.7/86.8\n', or better, a proper python list

Yes, it's possible to do region = [df.x.min(), df.x.max(), df.y.min(), df.y.max()], but gmtinfo -I does rounding (and I believe it rounds outside of the tight region?), and it has other smart functionality too.

It shouldn't be too hard to make pygmt.info support non-filename types, just need to use lib.virtualfile_from_vectors or the like.

Note that the return values of pygmt.info should be discussed at #147. This issue only tackles the input format type. Perhaps we should handle #147 before this tackling this issue.

Are you willing to help implement and maintain this feature? Yes