def test_facet_grid_scatter_plot(self):
## Check if it works without any error.
np.random.seed(1)
df = load_iris(target="species")
df["grid_col"] = np.random.choice(["c1", "c2"],
size=df.shape[0],
replace=True)
df["grid_row"] = np.random.choice(["r1", "r2", "r3"],
size=df.shape[0],
replace=True)
facet_grid_scatter_plot(data=df,
row="grid_row",
col="grid_col",
x="petal_width",
y="petal_length",
c="sepal_width")
facet_grid_scatter_plot(data=df,
row="grid_row",
col="grid_col",
x="petal_width",
y="petal_length",
hue="species")
def test_load_iris(self):
"""unittest for load_iris"""
target = "species"
cats = {"setosa", "versicolor", "virginica"}
df = load_iris(target=target)
self.assertTrue(isinstance(df, pd.DataFrame))
self.assertEqual((150, 5), df.shape)
self.assertEqual(target, list(df.columns)[-1])
self.assertEqual(cats, set(df[target].unique()))
def setUpClass(cls) -> None:
np.random.seed(1)
## iris (multi-class classification)
df = load_iris(target="label")
df = df.sample(frac=1, replace=False)
cls.iris_X = df.drop("label", axis=1)
cls.iris_y = df["label"]
cls.iris_plr = GridSearchCV(LogisticRegression(solver="liblinear",
multi_class="auto"),
param_grid={"C": [0.1, 1]},
cv=3,
iid=False,
return_train_score=True)
cls.iris_plr.fit(cls.iris_X, cls.iris_y)
## boston dataset (regression)
target = "price"
df = load_boston(target=target)
df = df.sample(frac=1, replace=False)
cls.boston_X = df.drop("price", axis=1)
cls.boston_y = df["price"]
cls.boston_tree = GridSearchCV(DecisionTreeRegressor(),
param_grid={"max_depth": [3, 5, 7]},
cv=5,
scoring="neg_mean_squared_error",
iid=False,
return_train_score=False)
cls.boston_tree.fit(cls.boston_X, cls.boston_y)
cls.boston_enet = GridSearchCV(ElasticNet(normalize=True),
param_grid={
"alpha": [0.1, 1.0],
"l1_ratio": [0.1, 0.5, 0.9]
},
cv=3,
scoring="neg_mean_squared_error",
iid=False,
return_train_score=True)
cls.boston_enet.fit(cls.boston_X, cls.boston_y)
## breast (binary classification)
df = load_breast_cancer("label")
cls.breast_cancer_X = df.drop("label", axis=1)
cls.breast_cancer_y = df["label"]
cls.breast_cancer_plr = GridSearchCV(
LogisticRegression(solver="liblinear"),
param_grid={"C": [0.1, 1]},
cv=3)
cls.breast_cancer_plr.fit(cls.breast_cancer_X, cls.breast_cancer_y)
def test_to_excel(self):
df_iris = load_iris()
df_boston = load_boston()
df_cancer = load_breast_cancer().loc[:10, :]
from openpyxl import load_workbook
sheets = ["iris", "boston"]
with TempDir() as temp_dir:
excel_file = temp_dir.joinpath("new_file.xlsx")
## Case 1) Create a new file
to_excel(df=df_iris,
file=excel_file,
sheet=sheets[0],
libreoffice=True)
wb1 = load_workbook(str(excel_file))
self.assertEqual([sheets[0]], wb1.sheetnames)
## whether we can read the excel through pandas.read_excel
dg1 = pd.read_excel(excel_file, sheet_name=sheets[0])
self.assertEqual(df_iris.columns.tolist(), dg1.columns.tolist())
self.assertEqual(df_iris.shape, dg1.shape)
## Case 2) Add a sheet to an existing excel file
to_excel(df=df_boston,
file=excel_file,
sheet=sheets[1],
libreoffice=True)
wb2 = load_workbook(str(excel_file))
self.assertEqual(sheets, wb2.sheetnames)
dg2 = pd.read_excel(excel_file, sheet_name=sheets[1])
self.assertEqual(df_boston.columns.tolist(), dg2.columns.tolist())
self.assertEqual(df_boston.shape, dg2.shape)
## Case 3) Overwrite an existing sheet with a new table
## This table has no style.
to_excel(df=df_cancer,
file=excel_file,
sheet=sheets[0],
libreoffice=False)
wb3 = load_workbook(str(excel_file))
self.assertEqual(sheets, wb3.sheetnames)
dg3 = pd.read_excel(excel_file, sheet_name=sheets[0])
self.assertEqual(df_cancer.columns.tolist(), dg3.columns.tolist())
self.assertEqual(df_cancer.shape, dg3.shape)
def test_bins_heatmap(self):
# check if this function works without error
np.random.seed(2)
df = load_iris(target="species")
df["cat1"] = np.random.choice(["a", "b"],
size=df.shape[0],
replace=True)
bins_heatmap(df,
cat1="cat1",
cat2="species",
x="petal_width",
y="petal_length",
target="sepal_length",
n_bins=3)
def test_bins_by_tree(self):
"""unittest for bins_by_tree"""
df = load_iris(target="species")
cols = list(df.columns)
n_bins = 3
## Case) The target variable is continuous
bins = bins_by_tree(df,
field=cols[2],
target=cols[3],
target_is_continuous=True,
n_bins=n_bins,
n_points=200,
precision=1)
cats = sorted(bins.unique())
self.assertIsInstance(bins, pd.Series)
self.assertEqual(len(bins.unique()), n_bins)
for cat in cats:
self.assertIsInstance(cat, pd.Interval)
self.assertEqual(cat.closed, "right")
self.assertEqual(cats[0].left, -np.inf)
self.assertEqual(cats[-1].right, np.inf)
## Case) The target variable is not continuous
bins = bins_by_tree(df,
field=cols[2],
target=cols[-1],
target_is_continuous=False,
n_bins=n_bins,
n_points=200,
precision=1)
cats = sorted(bins.unique())
self.assertIsInstance(bins, pd.Series)
self.assertEqual(len(bins.unique()), n_bins)
for cat in cats:
self.assertIsInstance(cat, pd.Interval)
self.assertEqual(cat.closed, "right")
self.assertEqual(cats[0].left, -np.inf)
self.assertEqual(cats[-1].right, np.inf)
def test_visualize_two_fields(self):
# check if the function works without any error
np.random.seed(1)
df = load_iris(target="species")
df["cat"] = np.random.choice(["a","b","c"],
size=df.shape[0],
replace=True)
inspector = Inspector(df)
## continuous x continuous
inspector.visualize_two_fields("sepal_width","sepal_length")
## continuous x categorical
inspector.visualize_two_fields("sepal_length", "species")
## categorical x continuous
inspector.visualize_two_fields("species", "petal_width")
## categorical x categorical
inspector.visualize_two_fields("species","cat")
try:
from adhoc.processing import Inspector
from adhoc.modeling import show_tree
from adhoc.utilities import load_iris, facet_grid_scatter_plot, bins_heatmap
except ImportError:
import sys
sys.path.append("..")
from adhoc.processing import Inspector
from adhoc.modeling import show_tree
from adhoc.utilities import load_iris, facet_grid_scatter_plot, bins_heatmap
# +
np.random.seed(1)
df = load_iris(target="species")
df["cat1"] = np.random.choice(["a", "b", "c"], size=df.shape[0], replace=True)
df["cat2"] = (df.iloc[:, 0] * df.iloc[:, 1] - df.iloc[:, 2] * df.iloc[:, 3] >
11).map({
True: 1,
False: 0
})
inspector = Inspector(df)
inspector ## 4 continuous variables and 3 categorical variables
# -
inspector.visualize_two_fields("sepal_width",
"sepal_length") ## continuous x continuous
inspector.visualize_two_fields("petal_width",