Enhancements to range handling of Histograms in Data Explorer

While testing #9482, I did some regression testing around histograms' display in data explorer, got together with @wesm, and identified some issues that would be worth improving.

Generate a data frame using the following Python script

import pandas as pd
import numpy as np

data = {
    "pure_small_integers": [0, 1, 2, 3, 4],
    "negative_integers": [-5, -4, -3, -2, -2],
    "floats_with_decimals": [0.5, 1.5, 2.5, 2.5, 2.5],
    "floats_that_are_whole": [0.0, 1.0, 2.0, 2.0, 2.0],
    "mixed_ints_and_decimals": [0, 1.25, 2, 3.5, 3.5],
    "single_value_integer": [7, 7, 7, 7, 7],
    "single_value_decimal": [3.3, 3.3, 3.3, 3.3, 3.3],
    "integers_with_nans": [0, 1, np.nan, 3, 3],
    "decimals_with_nans": [0.5, np.nan, 2.5, 2.5, 2.5],
    "very_large_integers": [1000000, 1000001, 1000002, 1000003, 1000005],
    "tiny_decimals": [0.001, 0.002, 0.003, 0.002, 0.003],
    "negative_decimals": [-2.25, -1.75, -2.0, -2.0, -1.75],
    "booleans": [True, False, True, False, True],
    "mixed_numeric_strings": ["1", "2", "3.5", "2", "3.5"],
    "mixed_object": [0, "1", 2, "3.5", 4],
    "inf_nan_present": [1.0, np.inf, -np.inf, np.nan, 3.5]
}

df_test_cases = pd.DataFrame(data)

df_test_cases.head()

Sub-issue A

• In pure_small_integers, range 2 is in both lower and upper bounds (0-2 and 2-4).
• It should be 0-1 and 2-4

Sub-issue B

• Evaluating floats_with_decimals, floats_that_are_whole, and mixed_ints_and_decimals, please note that the number of decimal places is now inconsistent.
• We should consider using same value formatting logic that is in the data grid, or have a logic that is different but at least consistent

Sub-issue C

• It is confusing when range is inclusive or exclusive of extreme bounds.
• It would be helpful to have () or [] added to the ranges to make it clear to the user if range is exclusive or inclusive.

Sub-issue D

• When evaluating mixed_object, notice that no histogram is generated in the data frame tab.
• We should consider enabling handling of object arrays so that sparklines can be computed in these cases.

Sub-issue E

• Posted as part of description by @hadley at #9833

(This display also reveals another challenge of histograms here: if you have one more unique value than number of bins, one of the bins is much larger than the others)