def test_seven_unique(self):
from dataworkspaces.kits.jupyter import _metric_col_to_colormap
from pandas.testing import assert_series_equal
bins = _metric_col_to_colormap(
pandas.Series([0.2, 1.4, numpy.nan, 1.2, 1.0, 1.0, 0.8, 0.4, 1.5]))
assert_series_equal(pandas.Series([1, 5, -1, 4, 3, 3, 2, 1, 5]),
bins,
check_dtype=False)
def test_combined_bins(self):
""""Test case from real bug where qcut() returns fewer bins than we asked"""
# there are 5 unique values, but qcut() will put it into 4 bins.
from dataworkspaces.kits.jupyter import _metric_col_to_colormap
from pandas.testing import assert_series_equal
data = pandas.Series([numpy.nan, 0.729885, 0.655172, 0.729885, numpy.nan, 0.729885, 0.747126, 0.729885, 0.729885, 0.701149, 0.729885, 0.758621])
bins = _metric_col_to_colormap(data)
expected=pandas.Series([-1,2,1,2,-1,2,5,2,2,1,2,5])
assert_series_equal(expected, bins, check_dtype=False)
def test_three_unique(self):
from dataworkspaces.kits.jupyter import _metric_col_to_colormap
from pandas.testing import assert_series_equal
bins = _metric_col_to_colormap(pandas.Series([1.4, numpy.nan, 1.2, 1.0]))
assert_series_equal(pandas.Series([4,-1,3, 2]), bins, check_dtype=False)
def test_two_unique(self):
from dataworkspaces.kits.jupyter import _metric_col_to_colormap
from pandas.testing import assert_series_equal
bins = _metric_col_to_colormap(pandas.Series([1.4, numpy.nan, 1.2]))
assert_series_equal(pandas.Series([4,-1,2]), bins)
def test_random(self):
from dataworkspaces.kits.jupyter import _metric_col_to_colormap
import random
random.seed(1)
data = pandas.Series([random.gauss(5, 1) for i in range(100)])
bins = _metric_col_to_colormap(data)
