fix documentation typos in docstrings in transforms.py, and in transf…

…orm_bivariate sphinx doc, and adjust testing accordingly to restructuring

docs/user_guide/percentage_change.rst

@@ -32,7 +32,7 @@ The first step is to import the required function from the distrx package.
 
 .. code-block:: python
 
-    from distrx import transform_univariate
+    from distrx import transform_bivariate
 
 Different transformation functions can be chosen through specifying a string parameter of which
 transform you would like to apply to your data. In this case, it is the following.
@@ -47,4 +47,5 @@ transform you would like to apply to your data. In this case, it is the followin
 
 ``mu_tx`` and ``sigma_tx`` are simply the percentage change for each county and their corresponding
 standard errors, respectively. ``sigma_tx`` has already been scaled the appropriate sample size so
-we **should not** scale it additionally to obtain a confidence interval.
+we **should not** scale it additionally with some function of othe sample size to obtain a
+confidence interval.

src/distrx/transforms.py

@@ -62,7 +62,7 @@ def log_trans(
 
         .. math::
 
-            \log(mu), \frac{\sigma}{\mu} \cdot \frac{1}{\sqrt{n}}
+            \\log(mu), \\frac{\\sigma}{\\mu} \\cdot \\frac{1}{\\sqrt{n}}
 
         Parameters
         ----------
@@ -86,7 +86,7 @@ def logit_trans(
 
         .. math::
 
-            \log(\frac{\mu}{1 - \mu}), \frac{\sigma}{\mu \cdot (1 - \mu)} \cdot \frac{1}{\sqrt{n}}
+            \\log(\\frac{\\mu}{1 - \\mu}), \\frac{\\sigma}{\\mu \\cdot (1 - \\mu)} \\cdot \\frac{1}{\\sqrt{n}}
 
         Parameters
         ----------
@@ -110,7 +110,7 @@ def exp_trans(
 
         .. math::
 
-            \exp(\mu), \sigma \cdot \exp(\mu) \cdot \frac{1}{\sqrt{n}}
+            \\exp(\\mu), \\sigma \\cdot \\exp(\\mu) \\cdot \\frac{1}{\\sqrt{n}}
 
         Parameters
         ----------
@@ -134,7 +134,7 @@ def expit_trans(
 
         .. math::
 
-            \frac{1}{1 + \exp(-\mu)}, \sigma \cdot \frac{\exp(\mu)}{(1 + \exp(\mu))^2} \cdot \frac{1}{\sqrt{n}}
+            \\frac{1}{1 + \\exp(-\\mu)}, \\sigma \\cdot \\frac{\\exp(\\mu)}{(1 + \\exp(\\mu))^2} \\cdot \\frac{1}{\\sqrt{n}}
 
         Parameters
         ----------
@@ -192,7 +192,7 @@ def percentage_change_trans(
 
         .. math::
 
-            \frac{p_y}{p_x} - 1, \sqrt{\frac{\sigma_y^2}{n_y\mu_x^2} + \frac{\mu_y^2\sigma_x^2}{n_x\mu_x^4}}
+            \\frac{p_y}{p_x} - 1, \\sqrt{\\frac{\\sigma_y^2}{n_y\\mu_x^2} + \\frac{\\mu_y^2\\sigma_x^2}{n_x\\mu_x^4}}
 
         Parameters
         ----------

tests/test_transforms.py

@@ -4,14 +4,9 @@
 import pandas as pd
 import pytest
 
-from distrx.transforms import (
-    delta_method,
-    transform_data,
-    transform_percentage_change,
-    transform_percentage_change_experiment,
-)
-
-TRANSFORM_DICT = {
+from distrx.transforms import transform_bivariate, transform_univariate
+
+UNIVARIATE_TRANSFORM_DICT = {
     "log": [np.log, lambda x: 1.0 / x],
     "logit": [lambda x: np.log(x / (1.0 - x)), lambda x: 1.0 / (x * (1.0 - x))],
     "exp": [np.exp, np.exp],
@@ -20,96 +15,98 @@
         lambda x: np.exp(-x) / (1.0 + np.exp(-x)) ** 2,
     ],
 }
-TRANSFORM_LIST = list(TRANSFORM_DICT.keys())
-FUNCTION_LIST = [transform_data, delta_method]
+UNIVARIATE_TRANSFORM_LIST = list(UNIVARIATE_TRANSFORM_DICT.keys())
+
+# BIVARIATE_TRANSFORM_DICT = {
+#     "percentage change": [np.]
+# }
+BIVARIATE_TRANSFORM_LIST = ["percentage_change"]
+# FUNCTION_LIST = [transform_data, delta_method]
 VALS = [0.1] * 2
+N = len(VALS)
 
 
-@pytest.mark.parametrize("transform", TRANSFORM_LIST)
+@pytest.mark.parametrize("transform", UNIVARIATE_TRANSFORM_LIST)
 def test_method_name_valid(transform):
     """Raise ValueError for invalue `method`."""
     with pytest.raises(ValueError):
-        transform_data(VALS, VALS, transform, method="dummy")
+        transform_univariate(VALS, VALS, N, transform, method="dummy")
 
 
-@pytest.mark.parametrize("function", FUNCTION_LIST)
-@pytest.mark.parametrize("transform", TRANSFORM_LIST)
-def test_input_len_match(function, transform):
+@pytest.mark.parametrize("transform", UNIVARIATE_TRANSFORM_LIST)
+def test_input_len_match(transform):
     """Raise ValueError if lengths of input vectors don't match."""
     with pytest.raises(ValueError):
-        function(VALS, VALS * 2, transform)
+        transform_univariate(VALS, VALS * 2, N, transform)
 
 
-@pytest.mark.parametrize("function", FUNCTION_LIST)
-@pytest.mark.parametrize("transform", TRANSFORM_LIST)
-def test_sigma_negative(function, transform):
+@pytest.mark.parametrize("transform", UNIVARIATE_TRANSFORM_LIST)
+def test_sigma_negative(transform):
     """Raise ValueError if `sigma` contains negative values."""
     vals = VALS + [-0.1]
     with pytest.raises(ValueError):
-        function(vals, vals, transform)
+        transform_univariate(vals, vals, N, transform)
 
 
-@pytest.mark.parametrize("function", FUNCTION_LIST)
-@pytest.mark.parametrize("transform", TRANSFORM_LIST)
-def test_sigma_zero(function, transform):
+@pytest.mark.parametrize("transform", UNIVARIATE_TRANSFORM_LIST)
+def test_sigma_zero(transform):
     """Display warning if `sigma` contains zeros."""
     vals = VALS + [0.0]
     with pytest.warns(UserWarning):
-        function(vals, vals, transform)
+        transform_univariate(vals, vals, N, transform)
 
 
-@pytest.mark.parametrize("function", FUNCTION_LIST)
-def test_transform_name_valid(function):
+def test_transform_name_valid():
     """Raise ValueError for invalid `transform`."""
     with pytest.raises(ValueError):
-        function(VALS, VALS, "dummy")
+        transform_univariate(VALS, VALS, N, "dummy")
+        transform_bivariate(VALS, VALS, VALS, VALS, "dummy")
 
 
-@pytest.mark.parametrize("function", FUNCTION_LIST)
-@pytest.mark.parametrize("transform", TRANSFORM_LIST)
-def test_output_type(function, transform):
+@pytest.mark.parametrize("transform", UNIVARIATE_TRANSFORM_LIST)
+def test_output_type(transform):
     """Output should be numpy arrays."""
-    mu, sigma = function(VALS, VALS, transform)
+    mu, sigma = transform_univariate(VALS, VALS, N, transform)
     assert isinstance(mu, np.ndarray)
     assert isinstance(sigma, np.ndarray)
 
 
-@pytest.mark.parametrize("function", FUNCTION_LIST)
-@pytest.mark.parametrize("transform", TRANSFORM_LIST)
-def test_outout_len_match(function, transform):
+@pytest.mark.parametrize("transform", UNIVARIATE_TRANSFORM_LIST)
+def test_outout_len_match(transform):
     """Length of output vectors should match."""
-    mu, sigma = function(VALS, VALS, transform)
+    mu, sigma = transform_univariate(VALS, VALS, N, transform)
     assert len(mu) == len(sigma)
 
 
-@pytest.mark.parametrize("transform", TRANSFORM_LIST)
+@pytest.mark.parametrize("transform", UNIVARIATE_TRANSFORM_LIST)
 def test_delta_result(transform):
     """Check expected results."""
     mu = np.random.uniform(0.1, 1.0, size=10)
     sigma = np.random.uniform(0.1, 1.0, size=10)
-    mu_ref = TRANSFORM_DICT[transform][0](mu)
-    sigma_ref = sigma * TRANSFORM_DICT[transform][1](mu)
-    mu_trans, sigma_trans = delta_method(mu, sigma, transform)
+    mu_ref = UNIVARIATE_TRANSFORM_DICT[transform][0](mu)
+    sigma_ref = sigma * UNIVARIATE_TRANSFORM_DICT[transform][1](mu)
+    mu_trans, sigma_trans = transform_univariate(mu, sigma, N, transform)
     assert np.allclose(mu_trans, mu_ref)
     assert np.allclose(sigma_trans, sigma_ref)
 
 
-def test_percentage_change():
-    x = np.random.normal(1, 0.1, 1000)
-    y = np.random.normal(1.1, 0.1, 1000)
-    z = np.random.normal(1, 0.1, 1001)
-    p, sigma = transform_percentage_change_experiment(x, y)
-    assert 0 < p and p < 1
-    assert 0 < sigma and sigma < 1
-    with pytest.raises(ValueError):
-        transform_percentage_change_experiment(x, z)
+# TODO: DEPRECATE
+# def test_percentage_change():
+#     x = np.random.normal(1, 0.1, 1000)
+#     y = np.random.normal(1.1, 0.1, 1000)
+#     z = np.random.normal(1, 0.1, 1001)
+#     p, sigma = transform_percentage_change_experiment(x, y)
+#     assert 0 < p and p < 1
+#     assert 0 < sigma and sigma < 1
+#     with pytest.raises(ValueError):
+#         transform_percentage_change_experiment(x, z)
 
 
 def test_percentage_change_counts():
     x = np.random.choice([0, 1], size=1000, p=[0.1, 0.9])
     y = np.random.choice([0, 1], size=1100, p=[0.2, 0.8])
-    mu, sigma = transform_percentage_change(
-        (x == 1).sum(), len(x), (y == 1).sum(), len(y)
+    mu, sigma = transform_bivariate(
+        (x == 1).sum(), len(x), (y == 1).sum(), len(y), "percentage_change"
     )
     assert -1 <= mu and mu < np.inf
     assert 0 < sigma and sigma < 1
@@ -120,17 +117,19 @@ def test_percentage_change_input():
     c_x, n_x = 100, 1000
     c_y, n_y = 200, 1050
     # with pytest.raises(TypeError):
-    transform_percentage_change(c_x, n_x, c_y, n_y)
+    transform_bivariate(c_x, n_x, c_y, n_y, "percentage_change")
 
     # base list input
     c_x = [100, 200]
     n_x = [1000, 1000]
     c_y = [300, 400]
     n_y = [1050, 1050]
     # with pytest.raises(TypeError):
-    transform_percentage_change(c_x, n_x, c_y, n_y)
+    transform_bivariate(c_x, n_x, c_y, n_y, "percentage_change")
 
     # dataframe input
     df = pd.DataFrame({"c_x": c_x, "n_x": n_x, "c_y": c_y, "n_y": n_y})
     # with pytest.raises(TypeError):
-    transform_percentage_change(df["c_x"], df["n_x"], df["c_y"], df["n_y"])
+    transform_bivariate(
+        df["c_x"], df["n_x"], df["c_y"], df["n_y"], "percentage_change"
+    )