Method test01

tables/tests/test_indexvalues.py:3458–3515

(self)

    open_mode = "w"

    def test01(self):
        np.random.seed(2)
        n = 700_000
        cs = 50_000
        nchunks = n // cs

        arr = np.zeros(
            (n,), dtype=[("index", "i8"), ("o", "i8"), ("value", "f8")]
        )
        arr["index"] = np.arange(n)
        arr["o"] = np.random.randint(-20_000, -15_000, size=n)
        arr["value"] = np.random.randn(n)

        node = self.h5file.create_group("/", "foo")
        table = self.h5file.create_table(
            node,
            "table",
            dict(
                index=tb.Int64Col(),
                o=tb.Int64Col(),
                value=tb.FloatCol(shape=()),
            ),
            expectedrows=10_000_000,
        )

        table.append(arr)

        self._reopen("a")

        v1 = np.unique(arr["o"])[0]
        v2 = np.unique(arr["o"])[1]
        res = np.array([v1, v2])
        selector = f"((o == {v1}) | (o == {v2}))"
        if common.verbose:
            print("selecting values: %s" % selector)

        table = self.h5file.root.foo.table

        result = np.unique(table.read_where(selector)["o"])
        np.testing.assert_almost_equal(result, res)
        if common.verbose:
            print("select entire table:")
            print(f"result: {result}\texpected: {res}")

        if common.verbose:
            print("index the column o")
        table.cols.o.create_index()  # this was triggering the issue

        if common.verbose:
            print("select via chunks")
        for i in range(nchunks):
            result = table.read_where(
                selector, start=i * cs, stop=(i + 1) * cs
            )
            result = np.unique(result["o"])
            np.testing.assert_almost_equal(np.unique(result), res)
            if common.verbose:
                print(f"result: {result}\texpected: {res}")