Cleared all the TODOs

Author: Jacobluke-

CHANGELOG.md

@@ -0,0 +1,14 @@
+# Release notes
+
+<!-- do not remove -->
+
+## 2023.03.29
+
+### New Features
+- Add new form of paired proportion plots for a better support of Repeated Measures
+
+
+## 0.2.3
+
+### Bug Fixes
+- Fixes a bug that jammed up when the xvar column was already a pandas Categorical. Now we check for this and act appropriately.

dabest/_bootstrap_tools.py

@@ -108,7 +108,13 @@ def __init__(
                 ttest_single = "NIL"
                 ttest_2_ind = "NIL"
                 ttest_2_paired = ttest_rel(x1, x2)[1]
-                wilcoxonresult = wilcoxon(x1, x2)[1]
+
+                try:
+                    wilcoxonresult = wilcoxon(x1, x2)[1]
+                except ValueError as e:
+                    warnings.warn("Wilcoxon test could not be performed. This might be due "
+                      "to identical values under the same group. "
+                      "Details: {}".format(e))
             mannwhitneyresult = "NIL"
 
             # Turns data into array, then tuple.

dabest/_dabest_object.py

@@ -442,26 +442,45 @@ def _check_errors(self, x, y, idx, experiment, experiment_label, x1_level):
                 raise ValueError(err0)
 
             # Check if the columns stated are valid
-            # TODO instead of traversing twice idx you can traverse only once
-            # and break the loop if the condition is not satisfied?
-            # TODO What if the type is not str and not tuple,list? missing raise Error
-            if all([isinstance(i, str) for i in idx]):
-                if len(pd.unique([t for t in idx]).tolist()) != 2:
+            # Initialize a flag to track if any element in idx is neither str nor (tuple, list)
+            valid_types = True
+
+            # Initialize variables to track the conditions for str and (tuple, list)
+            is_str_condition_met, is_tuple_list_condition_met = False, False
+
+            # Single traversal for optimization
+            for item in idx:
+                if isinstance(item, str):
+                    is_str_condition_met = True
+                elif isinstance(item, (tuple, list)) and len(item) == 2:
+                    is_tuple_list_condition_met = True
+                else:
+                    valid_types = False
+                    break  # Exit the loop if an invalid type is found
+
+            # Check if all types are valid
+            if not valid_types:
+                raise TypeError("Invalid type found in idx. Expected str, tuple, or list.")
+
+            # Handling str type condition
+            if is_str_condition_met:
+                if len(pd.unique(idx).tolist()) != 2:
                     err0 = "`mini_meta` is True, but `idx` ({})".format(idx)
-                    err1 = "does not contain exactly 2 columns."
+                    err1 = "does not contain exactly 2 unique columns."
                     raise ValueError(err0 + err1)
 
-            if all([isinstance(i, (tuple, list)) for i in idx]):
+            # Handling (tuple, list) type condition
+            if is_tuple_list_condition_met:
                 all_idx_lengths = [len(t) for t in idx]
                 if (array(all_idx_lengths) != 2).any():
-                    err1 = "`mini_meta` is True, but some idx "
-                    err2 = "in {} does not consist only of two groups.".format(idx)
+                    err1 = "`mini_meta` is True, but some elements in idx "
+                    err2 = "in {} do not consist only of two groups.".format(idx)
                     raise ValueError(err1 + err2)
 
-        # TODO can you have True mini_meta and delta2 at the same time?
+
         # Check if this is a 2x2 ANOVA case and x & y are valid columns
         # Create experiment_label and x1_level
-        if self.__delta2:
+        elif self.__delta2:
             if x is None:
                 error_msg = "If `delta2` is True. `x` parameter cannot be None. String or list expected"
                 raise ValueError(error_msg)
@@ -534,7 +553,6 @@ def _check_errors(self, x, y, idx, experiment, experiment_label, x1_level):
             else:
                 x1_level = self.__output_data[x[0]].unique()
 
-        # TODO what if experiment is None?
         elif experiment:
             experiment_label = self.__output_data[experiment].unique()
             x1_level = self.__output_data[x[0]].unique()
@@ -611,10 +629,7 @@ def _get_plot_data(self, x, y, all_plot_groups):
         # Added in v0.2.7.
         plot_data.dropna(axis=0, how="any", subset=[self.__yvar], inplace=True)
 
-        # TODO these comments should not be in the code but on the release notes of the package version
-        # Lines 131 to 140 added in v0.2.3.
-        # Fixes a bug that jammed up when the xvar column was already
-        # a pandas Categorical. Now we check for this and act appropriately.
+
         if isinstance(plot_data[self.__xvar].dtype, pd.CategoricalDtype):
             plot_data[self.__xvar].cat.remove_unused_categories(inplace=True)
             plot_data[self.__xvar].cat.reorder_categories(

dabest/_effsize_objects.py

@@ -299,7 +299,6 @@ def _compute_bca_intervals(self, sorted_bootstraps):
                 )
 
         else:
-            # TODO improve error handling, separate file with error messages?
             err1 = "The $lim_type limit of the BCa interval cannot be computed."
             err2 = "It is set to the effect size itself."
             err3 = "All bootstrap values were likely all the same."
@@ -330,9 +329,14 @@ def _perform_statistical_test(self):
 
         if self.__is_paired and not self.__proportional:
             # Wilcoxon, a non-parametric version of the paired T-test.
-            wilcoxon = spstats.wilcoxon(self.__control, self.__test)
-            self.__pvalue_wilcoxon = wilcoxon.pvalue
-            self.__statistic_wilcoxon = wilcoxon.statistic
+            try:
+                wilcoxon = spstats.wilcoxon(self.__control, self.__test)
+                self.__pvalue_wilcoxon = wilcoxon.pvalue
+                self.__statistic_wilcoxon = wilcoxon.statistic
+            except ValueError as e:
+                warnings.warn("Wilcoxon test could not be performed. This might be due "
+                  "to identical values under the same group. "
+                  "Details: {}".format(e))
 
             if self.__effect_size != "median_diff":
                 # Paired Student's t-test.
@@ -398,11 +402,10 @@ def _perform_statistical_test(self):
                 )
                 self.__pvalue_mann_whitney = mann_whitney.pvalue
                 self.__statistic_mann_whitney = mann_whitney.statistic
-            except ValueError:
-                # TODO At least print some warning?
-                # Occurs when the control and test are exactly identical
-                # in terms of rank (eg. all zeros.)
-                pass
+            except ValueError as e:
+                warnings.warn("Mann-Whitney test could not be performed. This might be due "
+                  "to identical rank values in both control and test groups. "
+                  "Details: {}".format(e))
 
             standardized_es = es.cohens_d(self.__control, self.__test, is_paired=None)
 
@@ -411,10 +414,9 @@ def _perform_statistical_test(self):
                 self.__proportional_difference = es.cohens_h(
                     self.__control, self.__test
                 )
-            except ValueError:
-                # TODO At least print some warning?
-                # Occur only when the data consists not only 0's and 1's.
-                pass
+            except ValueError as e:
+                warnings.warn(f"Calculation of Cohen's h failed. This method is applicable "
+                  f"only for binary data (0's and 1's). Details: {e}")
 
     def to_dict(self):
         """
@@ -567,87 +569,79 @@ def statistic_mcnemar(self):
 
     @property
     def pvalue_paired_students_t(self):
-        # TODO Missing docstring
         try:
             return self.__pvalue_paired_students_t
         except AttributeError:
             return npnan
 
     @property
     def statistic_paired_students_t(self):
-        # TODO Missing docstring
         try:
             return self.__statistic_paired_students_t
         except AttributeError:
             return npnan
 
     @property
     def pvalue_kruskal(self):
-        # TODO Missing docstring
         try:
             return self.__pvalue_kruskal
         except AttributeError:
             return npnan
 
     @property
     def statistic_kruskal(self):
-        # TODO Missing docstring
         try:
             return self.__statistic_kruskal
         except AttributeError:
             return npnan
 
     @property
     def pvalue_welch(self):
-        # TODO Missing docstring
         try:
             return self.__pvalue_welch
         except AttributeError:
             return npnan
 
     @property
     def statistic_welch(self):
-        # TODO Missing docstring
         try:
             return self.__statistic_welch
         except AttributeError:
             return npnan
 
     @property
     def pvalue_students_t(self):
-        # TODO Missing docstring
         try:
             return self.__pvalue_students_t
         except AttributeError:
             return npnan
 
     @property
     def statistic_students_t(self):
-        # TODO Missing docstring
         try:
             return self.__statistic_students_t
         except AttributeError:
             return npnan
 
     @property
     def pvalue_mann_whitney(self):
-        # TODO Missing docstring
         try:
             return self.__pvalue_mann_whitney
         except AttributeError:
             return npnan
 
     @property
     def statistic_mann_whitney(self):
-        # TODO Missing docstring
         try:
             return self.__statistic_mann_whitney
         except AttributeError:
             return npnan
 
     @property
     def pvalue_permutation(self):
-        # TODO Missing docstring
+        """
+        p value of permutation test
+        """
         return self.__PermutationTest_result.pvalue
 
     @property
@@ -663,12 +657,10 @@ def permutations(self):
 
     @property
     def permutations_var(self):
-         # TODO Missing docstring
         return self.__PermutationTest_result.permutations_var
 
     @property
     def proportional_difference(self):
-         # TODO Missing docstring
         try:
             return self.__proportional_difference
         except AttributeError:

dabest/_stats_tools/confint_1group.py

@@ -38,7 +38,9 @@ def compute_1group_jackknife(x, func, *args, **kwargs):
 
 
 def compute_1group_acceleration(jack_dist):
-    # TODO is it needed a function to just call one line?
+    """
+    Returns the accaleration value based on the jackknife distribution.
+    """
     from . import confint_2group_diff as ci_2g
 
     return ci_2g._calc_accel(jack_dist)

dabest/_stats_tools/confint_2group_diff.py

@@ -98,7 +98,9 @@ def compute_meandiff_jackknife(x0, x1, is_paired, effect_size):
 
 
 def _calc_accel(jack_dist):
-    # TODO Missing docstring
+    """
+    Given the Jackknife distribution, calculates the acceleration factor.
+    """
     jack_mean = npmean(jack_dist)
 
     numer = npsum((jack_mean - jack_dist) ** 3)

dabest/_stats_tools/effsize.py

@@ -210,7 +210,9 @@ def cohens_d(control:list|tuple|np.ndarray,
         M = np.mean(test) - np.mean(control)
         divisor = pooled_sd
 
-    # TODO what if divisor = 0?
+    if divisor == 0:
+        raise ValueError("The divisor is zero, indicating no variability in the data.")
+
     return M / divisor
 
 # %% ../../nbs/API/effsize.ipynb 8
@@ -312,8 +314,33 @@ def cliffs_delta(control:list|tuple|np.ndarray,
 
 # %% ../../nbs/API/effsize.ipynb 11
 def _compute_standardizers(control, test):
-    # TODO missing docstring
-    # For calculation of correlation; not currently used.
+    """
+    Computes the pooled and average standard deviations for two datasets.
+
+    This function is useful in the context of statistical analysis, particularly
+    when calculating standardized mean differences between two groups. It supports
+    both unpaired and paired data scenarios.
+
+    Parameters:
+    control (array-like): A numeric array representing the control group data.
+    test (array-like): A numeric array representing the test group data.
+
+    Returns:
+    tuple: A tuple containing two elements:
+        - pooled (float): The pooled standard deviation, calculated for unpaired two-group 
+                          scenarios. It is computed using the sample variances of the 
+                          control and test groups, weighted by their sample sizes.
+        - average (float): The average standard deviation, calculated for paired data 
+                           scenarios. It is the average of the sample standard deviations 
+                           of the control and test groups.
+
+    Note:
+    The function assumes that the input arrays are independent samples and calculates
+    the sample variances using N-1 degrees of freedom.
+
+    For calculation of correlation; not currently used.
+
+    """
     # from scipy.stats import pearsonr
 
     control_n = len(control)

dabest/misc_tools.py

@@ -33,7 +33,17 @@ def unpack_and_add(l, c):
 
 
 def print_greeting():
-    # TODO missing docstring
+    """
+    Generates a greeting message based on the current time, along with the version information of DABEST.
+
+    This function dynamically generates a greeting ('Good morning', 'Good afternoon', 'Good evening')
+    based on the current system time. It also retrieves and displays the version of DABEST (Data Analysis
+    using Bootstrap-Coupled ESTimation). The message includes a header with the DABEST version and the
+    current time formatted in a user-friendly manner.
+
+    Returns:
+    str: A formatted string containing the greeting message, DABEST version, and current time.
+    """
     from .__init__ import __version__
 
     line1 = "DABEST v{}".format(__version__)