The worldbank api doesn't list any data for Saint Barthélemy, BLM, an "overseas collectivity of France in the Caribbean"
with a population size of about 10 000 and an GDB of estimated 0.5 billion $.

Per se that is probably not a problem. But it implicates that one cannot create a LitPop Exposures object for Saint Barts anymore, since the pandas datareader dependency was dismissed and we started to read world bank data directly from world bank api
(PR #1033).

Unfortunately - or luckily - there is an integration test (climada.test.test_litpop_integr.TestLitPopExposure.test_BLM150_pass) checking for exactly this:

    def test_BLM150_pass(self):
        """Test from_countries for BLM at 150 arcsec, 2 data points"""
        ent = lp.LitPop.from_countries("BLM", res_arcsec=150, reference_year=2016)
        self.assertEqual(ent.gdf.shape[0], 2)

We can
1 - skip the test
2 - pick another country (don't think there is a point in that, as there is already the same test for Monaco)
3 - look for other data sources in case worldbank doesn't deliver

There is no comment in the test that says what is the point of it. Yet I kind of suspect it might be just about GDP from unlisted countries...

Suggestions are welcome. 😃 @peanutfun what is your opinion?