AutoStatLib 0.2.0__tar.gz → 0.2.2__tar.gz

{autostatlib-0.2.0/src/AutoStatLib.egg-info → autostatlib-0.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: AutoStatLib
-Version: 0.2.0
+Version: 0.2.2
 Summary: AutoStatLib - a simple statistical analysis tool
 Author: Stemonitis, SciWare LLC
 Author-email: konung-yaropolk <yaropolk1995@gmail.com>
@@ -531,6 +531,7 @@ License-File: LICENSE
 Requires-Dist: numpy
 Requires-Dist: scipy
 Requires-Dist: statsmodels
+Requires-Dist: pandas
 
 # AutoStatLib - python library for automated statistical analysis 
 
@@ -569,7 +570,7 @@ data_uniform = [list(np.random.uniform(i+3, i+1, n)) for i in range(groups)]
 
 
 # set the parameters:
-paired = False     # is groups dependend or not
+paired = False     # is groups dependent or not
 tails = 2          # two-tailed or one-tailed result
 popmean = 0        # population mean - only for single-sample tests needed
 
@@ -585,7 +586,7 @@ analysis.RunAuto()
 
 or you can choose specific tests:
 ```python
-# 2 groups independend:
+# 2 groups independent:
 analysis.RunTtest()
 analysis.RunMannWhitney()
 
@@ -594,10 +595,11 @@ analysis.RunTtestPaired()
 analysis.RunWilcoxon()
 
 # 3 and more independed groups comparison:
-analysis.RunAnova()
+analysis.RunOnewayAnova()
 analysis.RunKruskalWallis()
 
 # 3 and more depended groups comparison:
+analysis.RunOnewayAnovaRM()
 analysis.RunFriedman()
 
 # single group tests"
@@ -647,20 +649,40 @@ If errors occured, *GetResult()* returns an empty dictionary
 
 
 ---
-## Pre-Alpha dev status.
-
-### TODO:
-
---Kruskal-Wallis test - add Dunn's multiple comparisons
---Anova: add 2-way anova and 3-way(?)
-
-check:
---Wilcoxon signed-rank test and Mann-whitney - check mechanism of one-tailed calc, looks like it works wrong
-
-
-checked tests:
---Wilcoxon 2 tail - ok
---Mann-whitney 2 tail - ok
+## Pre-Alpha dev status.  
+  
+### TODO:  
+  
+-- Kruskal-Wallis test - add Dunn's multiple comparisons  
+-- Anova: add 2-way anova and 3-way anova  
+-- onevay Anova: add repeated measures (for normal dependent values) with and without Gaisser-Greenhouse correction  
+-- onevay Anova: add Brown-Forsithe and Welch (for normal independent values with unequal SDs between groups)  
+-- paired T-test: add ratio-paired t-test (ratios of paired values are consistent)  
+-- add Welch test (for norm data unequal variances)
+-- add Kolmogorov-smirnov test (unpaired nonparametric 2 sample, compare cumulative distributions)  
+-- add independent t-test with Welch correction (do not assume equal SDs in groups)  
+-- add correlation test, correlation diagram  
+-- add linear regression, regression diagram  
+-- add QQ plot
+-- n-sample tests: add onetail option
+  
+✅ done -- detailed normality test results
+
+  
+checked tests:  
+1-sample:  
+--Wilcoxon 2,1 tails - ok  
+--t-tests 2,1 tails -ok  
+  
+2-sample:  
+--Wilcoxon 2,1 tails - ok  
+--Mann-whitney 2,1 tails - ok  
+--t-tests 2,1 tails -ok  
+  
+n-sample:  
+--Kruskal-Wallis 2 tail - ok  
+--Friedman 2 tail - ok  
+--one-way ANOWA 2 tail - ok  
 
 
 

{autostatlib-0.2.0 → autostatlib-0.2.2}/README.md

@@ -35,7 +35,7 @@ data_uniform = [list(np.random.uniform(i+3, i+1, n)) for i in range(groups)]
 
 
 # set the parameters:
-paired = False     # is groups dependend or not
+paired = False     # is groups dependent or not
 tails = 2          # two-tailed or one-tailed result
 popmean = 0        # population mean - only for single-sample tests needed
 
@@ -51,7 +51,7 @@ analysis.RunAuto()
 
 or you can choose specific tests:
 ```python
-# 2 groups independend:
+# 2 groups independent:
 analysis.RunTtest()
 analysis.RunMannWhitney()
 
@@ -60,10 +60,11 @@ analysis.RunTtestPaired()
 analysis.RunWilcoxon()
 
 # 3 and more independed groups comparison:
-analysis.RunAnova()
+analysis.RunOnewayAnova()
 analysis.RunKruskalWallis()
 
 # 3 and more depended groups comparison:
+analysis.RunOnewayAnovaRM()
 analysis.RunFriedman()
 
 # single group tests"
@@ -113,20 +114,40 @@ If errors occured, *GetResult()* returns an empty dictionary
 
 
 ---
-## Pre-Alpha dev status.
-
-### TODO:
-
---Kruskal-Wallis test - add Dunn's multiple comparisons
---Anova: add 2-way anova and 3-way(?)
-
-check:
---Wilcoxon signed-rank test and Mann-whitney - check mechanism of one-tailed calc, looks like it works wrong
-
-
-checked tests:
---Wilcoxon 2 tail - ok
---Mann-whitney 2 tail - ok
+## Pre-Alpha dev status.  
+  
+### TODO:  
+  
+-- Kruskal-Wallis test - add Dunn's multiple comparisons  
+-- Anova: add 2-way anova and 3-way anova  
+-- onevay Anova: add repeated measures (for normal dependent values) with and without Gaisser-Greenhouse correction  
+-- onevay Anova: add Brown-Forsithe and Welch (for normal independent values with unequal SDs between groups)  
+-- paired T-test: add ratio-paired t-test (ratios of paired values are consistent)  
+-- add Welch test (for norm data unequal variances)
+-- add Kolmogorov-smirnov test (unpaired nonparametric 2 sample, compare cumulative distributions)  
+-- add independent t-test with Welch correction (do not assume equal SDs in groups)  
+-- add correlation test, correlation diagram  
+-- add linear regression, regression diagram  
+-- add QQ plot
+-- n-sample tests: add onetail option
+  
+✅ done -- detailed normality test results
+
+  
+checked tests:  
+1-sample:  
+--Wilcoxon 2,1 tails - ok  
+--t-tests 2,1 tails -ok  
+  
+2-sample:  
+--Wilcoxon 2,1 tails - ok  
+--Mann-whitney 2,1 tails - ok  
+--t-tests 2,1 tails -ok  
+  
+n-sample:  
+--Kruskal-Wallis 2 tail - ok  
+--Friedman 2 tail - ok  
+--one-way ANOWA 2 tail - ok  
 
 
 

autostatlib-0.2.0/src/AutoStatLib.egg-info/requires.txt → autostatlib-0.2.2/requirements.txt

@@ -1,3 +1,4 @@
 numpy
 scipy
 statsmodels
+pandas

{autostatlib-0.2.0 → autostatlib-0.2.2}/src/AutoStatLib/AutoStatLib.py

@@ -1,6 +1,8 @@
 import numpy as np
+import pandas as pd
 from statsmodels.stats.diagnostic import lilliefors
-from scipy.stats import ttest_rel, ttest_ind, ttest_1samp, wilcoxon, mannwhitneyu, f_oneway, kruskal, friedmanchisquare, shapiro, kstest, anderson, normaltest
+from statsmodels.stats.anova import AnovaRM
+from scipy.stats import ttest_rel, ttest_ind, ttest_1samp, wilcoxon, mannwhitneyu, f_oneway, kruskal, friedmanchisquare, shapiro, anderson, normaltest
 
 
 class __StatisticalTests():
@@ -8,37 +10,113 @@ class __StatisticalTests():
         Statistical tests mixin
     '''
 
-    def anova(self):
+    def run_test_auto(self):
+
+        if self.n_groups == 1:
+            if self.parametric:
+                self.run_test_by_id('t_test_single_sample')
+            else:
+                self.run_test_by_id('wilcoxon_single_sample')
+
+        elif self.n_groups == 2:
+            if self.paired:
+                if self.parametric:
+                    self.run_test_by_id('t_test_paired')
+                else:
+                    self.run_test_by_id('wilcoxon')
+            else:
+                if self.parametric:
+                    self.run_test_by_id('t_test_independent')
+                else:
+                    self.run_test_by_id('mann_whitney')
+
+        elif self.n_groups >= 3:
+            if self.paired:
+                if self.parametric:
+                    self.run_test_by_id('anova_1w_rm')
+                else:
+                    self.run_test_by_id('friedman')
+            else:
+                if self.parametric:
+                    self.run_test_by_id('anova_1w_ordinary')
+                else:
+                    self.run_test_by_id('kruskal_wallis')
+
+        else:
+            pass
+
+    def run_test_by_id(self, test_id):
+
+        test_names_dict = {
+            'anova_1w_ordinary': 'Ordinary One-Way ANOVA',
+            'anova_1w_rm': 'Repeated Measures One-Way ANOVA',
+            'friedman': 'Friedman test', 
+            'kruskal_wallis': 'Kruskal-Wallis test',
+            'mann_whitney': 'Mann-Whitney U test',
+            't_test_independent': 't-test for independent samples',
+            't_test_paired': 't-test for paired samples',
+            't_test_single_sample': 'Single-sample t-test',
+            'wilcoxon': 'Wilcoxon signed-rank test',
+            'wilcoxon_single_sample': 'Wilcoxon signed-rank test for single sample',
+        }
+
+        match test_id:
+            case 'anova_1w_ordinary': stat, p_value = self.anova_1w_ordinary()
+            case 'anova_1w_rm': stat, p_value = self.anova_1w_rm()
+            case 'friedman': stat, p_value = self.friedman()
+            case 'kruskal_wallis': stat, p_value = self.kruskal_wallis()
+            case 'mann_whitney': stat, p_value = self.mann_whitney()
+            case 't_test_independent': stat, p_value = self.t_test_independent()
+            case 't_test_paired': stat, p_value = self.t_test_paired()
+            case 't_test_single_sample': stat, p_value = self.t_test_single_sample()
+            case 'wilcoxon': stat, p_value = self.wilcoxon()
+            case 'wilcoxon_single_sample': stat, p_value = self.wilcoxon_single_sample()
+
+        if test_id in self.test_ids_dependent:
+            self.paired = True
+        else:
+            self.paired = False
+
+        self.test_name = test_names_dict[test_id]
+        self.test_id = test_id
+        self.test_stat = stat
+        self.p_value = p_value
+
+    def anova_1w_ordinary(self):
         stat, p_value = f_oneway(*self.data)
         self.tails = 2
         # if self.tails == 1 and p_value > 0.5:
         #     p_value /= 2
         # if self.tails == 1:
         #     p_value /= 2
-        self.test_name = 'ANOVA'
-        self.test_id = 'anova'
-        self.paired = False
-        self.test_stat = stat
-        self.p_value = p_value
+        return stat, p_value
+
+    def anova_1w_rm(self):
+        """
+        Perform repeated measures one-way ANOVA test.
+
+        Parameters:
+        data: list of lists, where each sublist represents repeated measures for a subject
+        """
+
+        df = self.matrix_to_dataframe(self.data)
+        res = AnovaRM(df, 'Value', 'Row', within=['Col']).fit()
+        stat = res.anova_table['F Value'][0]
+        p_value = res.anova_table['Pr > F'][0]
+
+        self.tails = 2
+        return stat, p_value
 
-    def friedman_test(self):
+    def friedman(self):
         stat, p_value = friedmanchisquare(*self.data)
         self.tails = 2
-        self.test_name = 'Friedman test'
-        self.test_id = 'friedman'
-        self.paired = True
-        self.test_stat = stat
-        self.p_value = p_value
+        return stat, p_value
 
-    def kruskal_wallis_test(self):
+    def kruskal_wallis(self):
         stat, p_value = kruskal(*self.data)
-        self.test_name = 'Kruskal-Wallis test'
-        self.test_id = 'kruskal_wallis'
-        self.paired = False
-        self.test_stat = stat
-        self.p_value = p_value
+        return stat, p_value
 
-    def mann_whitney_u_test(self):
+    def mann_whitney(self):
         stat, p_value = mannwhitneyu(
             self.data[0], self.data[1], alternative='two-sided')
         if self.tails == 1:
@@ -49,78 +127,53 @@ class __StatisticalTests():
         #     self.data[0], self.data[1], alternative='two-sided' if self.tails == 2 else 'less')
         # if self.tails == 1 and p_value > 0.5:
         #     p_value = 1-p_value
+        return stat, p_value
 
-        self.test_name = 'Mann-Whitney U test'
-        self.test_id = 'mann_whitney'
-        self.paired = False
-        self.test_stat = stat
-        self.p_value = p_value
-
-    def t_test_independend(self):
-        t_stat, t_p_value = ttest_ind(
+    def t_test_independent(self):
+        stat, p_value = ttest_ind(
             self.data[0], self.data[1])
         if self.tails == 1:
-            t_p_value /= 2
-        self.test_name = 't-test for independend samples'
-        self.test_id = 't_test_independend'
-        self.paired = False
-        self.test_stat = t_stat
-        self.p_value = t_p_value
+            p_value /= 2
+        return stat, p_value
 
     def t_test_paired(self):
-        t_stat, t_p_value = ttest_rel(
+        stat, p_value = ttest_rel(
             self.data[0], self.data[1])
         if self.tails == 1:
-            t_p_value /= 2
-        self.test_name = 't-test for paired samples'
-        self.test_id = 't_test_paired'
-        self.paired = True
-        self.test_stat = t_stat
-        self.p_value = t_p_value
+            p_value /= 2
+        return stat, p_value
 
     def t_test_single_sample(self):
         if self.popmean == None:
             self.popmean = 0
             self.AddWarning('no_pop_mean_set')
-        t_stat, t_p_value = ttest_1samp(self.data[0], self.popmean)
+        stat, p_value = ttest_1samp(self.data[0], self.popmean)
         if self.tails == 1:
-            t_p_value /= 2
-        self.test_name = 'Single-sample t-test'
-        self.test_id = 't_test_single_sample'
-        self.paired = False
-        self.test_stat = t_stat
-        self.p_value = t_p_value
+            p_value /= 2
+        return stat, p_value
 
+    def wilcoxon(self):
+        stat, p_value = wilcoxon(self.data[0], self.data[1])
+        if self.tails == 1:
+            p_value /= 2
+        return stat, p_value
+        
     def wilcoxon_single_sample(self):
         if self.popmean == None:
             self.popmean = 0
             self.AddWarning('no_pop_mean_set')
         data = [i - self.popmean for i in self.data[0]]
-        w_stat, p_value = wilcoxon(data)
+        stat, p_value = wilcoxon(data)
         if self.tails == 1:
             p_value /= 2
-        self.test_name = 'Wilcoxon signed-rank test for single sample'
-        self.test_id = 'wilcoxon_single_sample'
-        self.paired = False
-        self.test_stat = w_stat
-        self.p_value = p_value
-
-    def wilcoxon(self):
-        stat, p_value = wilcoxon(self.data[0], self.data[1])
-        if self.tails == 1:
-            p_value /= 2
-        self.test_name = 'Wilcoxon signed-rank test'
-        self.test_id = 'wilcoxon'
-        self.paired = True
-        self.test_stat = stat
-        self.p_value = p_value
+        return stat, p_value
 
 
 class __NormalityTests():
     '''
         Normality tests mixin
 
-        see the article about minimum sample size for tests:
+        see the article about minimal sample size for tests:
         Power comparisons of Shapiro-Wilk, Kolmogorov-Smirnov,
         Lilliefors and Anderson-Darling tests, Nornadiah Mohd Razali1, Yap Bee Wah1
     '''
@@ -171,7 +224,7 @@ class __NormalityTests():
 
     def anderson_get_p(self, data, dist='norm'):
         '''
-            calculating p-value for Anderson-Darling test using the method described here:            
+            calculating p-value for Anderson-Darling test using the method described here:
             Computation of Probability Associated with Anderson-Darling Statistic
             Lorentz Jantschi and Sorana D. Bolboaca, 2018 - Mathematics
 
@@ -199,6 +252,65 @@ class __NormalityTests():
         return ad, p
 
 
+class __Helpers():
+
+    def matrix_to_dataframe(self, matrix):
+        data = []
+        cols = []
+        rows = []
+
+        order_number = 1
+        for i, row in enumerate(matrix):
+            for j, value in enumerate(row):
+                data.append(value)
+                cols.append(i)
+                rows.append(j)
+                order_number += 1
+
+        df = pd.DataFrame(
+            {'Row': rows, 'Col': cols, 'Value': data})
+        return df
+
+    def create_results_dict(self) -> dict:
+
+        self.stars_int = self.make_stars()
+        self.stars_str = '*' * self.stars_int if self.stars_int else 'ns'
+
+        return {
+            'p-value': self.make_p_value_printed(),
+            'Significance(p<0.05)':  True if self.p_value.item() < 0.05 else False,
+            'Stars_Printed': self.stars_str,
+            'Test_Name': self.test_name,
+            'Groups_Compared': self.n_groups,
+            'Population_Mean': self.popmean if self.n_groups == 1 else 'N/A',
+            'Data_Normaly_Distributed': self.parametric,
+            'Parametric_Test_Applied': True if self.test_id in self.test_ids_parametric else False,
+            'Paired_Test_Applied': self.paired,
+            'Tails': self.tails,
+            'p-value_exact': self.p_value.item(),
+            'Stars':  self.stars_int,
+            # 'Stat_Value': self.test_stat.item(),
+            'Warnings': self.warnings,
+            'Groups_N': [len(self.data[i]) for i in range(len(self.data))],
+            'Groups_Median': [np.median(self.data[i]).item() for i in range(len(self.data))],
+            'Groups_Mean': [np.mean(self.data[i]).item() for i in range(len(self.data))],
+            'Groups_SD': [np.std(self.data[i]).item() for i in range(len(self.data))],
+            'Groups_SE': [np.std(self.data[i]).item() / np.sqrt(len(self.data)).item() for i in range(len(self.data))],
+            # actually returns list of lists of numpy dtypes of float64, next make it return regular floats:
+            'Samples': self.data,
+        }
+
+    def log(self, *args, **kwargs):
+        message = ' '.join(map(str, args))
+        # print(message, **kwargs)
+        self.summary += '\n' + message
+
+    def AddWarning(self, warning_id):
+        message = self.warning_ids_all[warning_id]
+        self.log(message)
+        self.warnings.append(message)
+
+
 class __TextFormatting():
     '''
         Text formatting mixin
@@ -293,45 +405,6 @@ class __TextFormatting():
             else:
                 self.log(i, ':', ' ' * shift, self.results[i])
 
-    def create_results_dict(self) -> dict:
-
-        self.stars_int = self.make_stars()
-        self.stars_str = '*' * self.stars_int if self.stars_int else 'ns'
-
-        return {
-            'p-value': self.make_p_value_printed(),
-            'Significance(p<0.05)':  True if self.p_value.item() < 0.05 else False,
-            'Stars_Printed': self.stars_str,
-            'Test_Name': self.test_name,
-            'Groups_Compared': self.n_groups,
-            'Population_Mean': self.popmean if self.n_groups == 1 else 'N/A',
-            'Data_Normaly_Distributed': self.parametric,
-            'Parametric_Test_Applied': True if self.test_id in self.test_ids_parametric else False,
-            'Paired_Test_Applied': self.paired,
-            'Tails': self.tails,
-            'p-value_exact': self.p_value.item(),
-            'Stars':  self.stars_int,
-            # 'Stat_Value': self.test_stat.item(),
-            'Warnings': self.warnings,
-            'Groups_N': [len(self.data[i]) for i in range(len(self.data))],
-            'Groups_Median': [np.median(self.data[i]).item() for i in range(len(self.data))],
-            'Groups_Mean': [np.mean(self.data[i]).item() for i in range(len(self.data))],
-            'Groups_SD': [np.std(self.data[i]).item() for i in range(len(self.data))],
-            'Groups_SE': [np.std(self.data[i]).item() / np.sqrt(len(self.data)).item() for i in range(len(self.data))],
-            # actually returns list of lists of numpy dtypes of float64, next make it return regular floats:
-            'Samples': self.data,
-        }
-
-    def log(self, *args, **kwargs):
-        message = ' '.join(map(str, args))
-        # print(message, **kwargs)
-        self.summary += '\n' + message
-
-    def AddWarning(self, warning_id):
-        message = self.warning_ids_all[warning_id]
-        self.log(message)
-        self.warnings.append(message)
-
 
 class __InputFormatting():
     def floatify_recursive(self, data):
@@ -349,7 +422,7 @@ class __InputFormatting():
                 return None
 
 
-class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting, __InputFormatting):
+class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting, __InputFormatting, __Helpers):
     '''
         The main class
         *documentation placeholder*
@@ -372,21 +445,49 @@ class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting
         self.n_groups = len(self.groups_list)
         self.warning_flag_non_numeric_data = False
         self.summary = ''
-        self.test_ids_parametric = ['anova',
-                                    't_test_independend',
-                                    't_test_paired',
-                                    't_test_single_sample',]
+
+        # test IDs classification:
         self.test_ids_all = [  # in aplhabetical order
-            'anova',
+            'anova_1w_ordinary',
+            'anova_1w_rm',
             'friedman',
             'kruskal_wallis',
             'mann_whitney',
-            't_test_independend',
+            't_test_independent',
             't_test_paired',
             't_test_single_sample',
             'wilcoxon',
             'wilcoxon_single_sample',
         ]
+        self.test_ids_parametric = [
+            'anova_1w_ordinary',
+            'anova_1w_rm'
+            't_test_independent',
+            't_test_paired',
+            't_test_single_sample',
+        ]
+        self.test_ids_dependent = [
+            'anova_1w_rm',
+            'friedman',
+            't_test_paired',
+            'wilcoxon',
+        ]
+        self.test_ids_3sample = [
+            'anova_1w_ordinary',
+            'anova_1w_rm',
+            'friedman',
+            'kruskal_wallis',
+        ]
+        self.test_ids_2sample = [
+            'mann_whitney',
+            't_test_independent',
+            't_test_paired',
+            'wilcoxon',
+        ]
+        self.test_ids_1sample = [
+            't_test_single_sample',
+            'wilcoxon_single_sample',
+        ]
         self.warning_ids_all = {
             # 'not-numeric':                     '\nWarning: Non-numeric data was found in input and ignored.\n         Make sure the input data is correct to get the correct results\n',
             'param_test_with_non-normal_data': '\nWarning: Parametric test was manualy chosen for Not-Normaly distributed data.\n         The results might be skewed. \n         Please, run non-parametric test or preform automatic test selection.\n',
@@ -425,28 +526,18 @@ class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting
             assert self.data, 'There is no input data'
             assert self.tails in [1, 2], 'Tails parameter can be 1 or 2 only'
             assert test in self.test_ids_all or test == 'auto', 'Wrong test id choosen, ensure you called correct function'
-            assert not (self.n_groups > 1
-                        and (test == 't_test_single_sample'
-                             or test == 'wilcoxon_single_sample')), 'Only one group of data must be given for single-group tests'
             assert all(len(
                 group) >= 4 for group in self.data), 'Each group must contain at least four values'
-            assert not (self.paired == True and not all(len(lst) == len(
-                self.data[0]) for lst in self.data)), 'Paired groups must be the same length'
-            assert not (test == 'friedman' and not all(len(lst) == len(
-                self.data[0]) for lst in self.data)), 'Paired groups must be the same length for Friedman Chi Square test'
-            assert not (test == 't_test_paired' and not all(len(lst) == len(
-                self.data[0]) for lst in self.data)), 'Paired groups must be the same length for Paired t-test'
-            assert not (test == 'wilcoxon' and not all(len(lst) == len(
-                self.data[0]) for lst in self.data)), 'Paired groups must be the same length for Wilcoxon signed-rank test'
-            assert not (test == 'friedman' and self.n_groups <
-                        3), 'At least three groups of data must be given for 3-groups tests'
-            assert not ((test == 'anova'
-                         or test == 'kruskal_wallis') and self.n_groups < 2), 'At least two groups of data must be given for ANOVA or Kruskal Wallis tests'
-            assert not ((test == 'wilcoxon'
-                         or test == 't_test_independend'
-                         or test == 't_test_paired'
-                         or test == 'mann_whitney')
-                        and self.n_groups != 2), 'Only two groups of data must be given for 2-groups tests'
+            assert not (self.paired == True
+                        and not all(len(lst) == len(self.data[0]) for lst in self.data)), 'Paired groups must have the same length'
+            assert not (test in self.test_ids_dependent
+                        and not all(len(lst) == len(self.data[0]) for lst in self.data)), 'Groups must have the same length for dependent groups test'
+            assert not (test in self.test_ids_2sample
+                        and self.n_groups != 2), f'Only two groups of data must be given for 2-groups tests, got {self.n_groups}'
+            assert not (test in self.test_ids_1sample
+                        and self.n_groups > 1), f'Only one group of data must be given for single-group tests, got {self.n_groups}'
+            assert not (test in self.test_ids_3sample
+                        and self.n_groups < 3), f'At least three groups of data must be given for multi-groups tests, got {self.n_groups}'
         except AssertionError as error:
             self.log('\nTest  :', test)
             self.log('Error :', error)
@@ -490,27 +581,13 @@ class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting
         if not test == 'auto' and self.parametric and not test in self.test_ids_parametric:
             self.AddWarning('non-param_test_with_normal_data')
 
-        if test == 'anova':
-            self.anova()
-        elif test == 'friedman':
-            self.friedman_test()
-        elif test == 'kruskal_wallis':
-            self.kruskal_wallis_test()
-        elif test == 'mann_whitney':
-            self.mann_whitney_u_test()
-        elif test == 't_test_independend':
-            self.t_test_independend()
-        elif test == 't_test_paired':
-            self.t_test_paired()
-        elif test == 't_test_single_sample':
-            self.t_test_single_sample()
-        elif test == 'wilcoxon':
-            self.wilcoxon()
-        elif test == 'wilcoxon_single_sample':
-            self.wilcoxon_single_sample()
+        # run the test
+
+        if test in self.test_ids_all:
+            self.run_test_by_id(test)
         else:
-            self.log('Automatic test selection preformed.')
-            self.__auto()
+            self.run_test_auto()
+
 
         # print the results
         self.results = self.create_results_dict()
@@ -523,32 +600,7 @@ class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting
         if self.verbose == True:
             print(self.summary)
 
-    def __auto(self):
 
-        if self.n_groups == 2:
-            if self.paired:
-                if self.parametric:
-                    return self.t_test_paired()
-                else:
-                    return self.wilcoxon()
-            else:
-                if self.parametric:
-                    return self.t_test_independend()
-                else:
-                    return self.mann_whitney_u_test()
-        elif self.n_groups == 1:
-            if self.parametric:
-                return self.t_test_single_sample()
-            else:
-                return self.wilcoxon_single_sample()
-        else:
-            if self.paired:
-                return self.friedman_test()
-            else:
-                if self.parametric:
-                    return self.anova()
-                else:
-                    return self.kruskal_wallis_test()
 
     # public methods:
     def RunAuto(self):
@@ -557,8 +609,11 @@ class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting
     def RunManual(self, test):
         self.__run_test(test)
 
-    def RunAnova(self):
-        self.__run_test(test='anova')
+    def RunOnewayAnova(self):
+        self.__run_test(test='anova_1w_ordinary')
+
+    def RunOnewayAnovaRM(self):
+        self.__run_test(test='anova_1w_rm')
 
     def RunFriedman(self):
         self.__run_test(test='friedman')
@@ -570,7 +625,7 @@ class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting
         self.__run_test(test='mann_whitney')
 
     def RunTtest(self):
-        self.__run_test(test='t_test_independend')
+        self.__run_test(test='t_test_independent')
 
     def RunTtestPaired(self):
         self.__run_test(test='t_test_paired')
@@ -603,6 +658,9 @@ class StatisticalAnalysis(__StatisticalTests, __NormalityTests, __TextFormatting
         else:
             return self.summary
 
+    def GetTestIDs(self):
+        return self.test_ids_all
+
     def PrintSummary(self):
         print(self.summary)
 

{autostatlib-0.2.0 → autostatlib-0.2.2}/src/AutoStatLib/_version.py

@@ -1,2 +1,2 @@
 # AutoStatLib package version:
-__version__ = "0.2.0"
+__version__ = "0.2.2"

{autostatlib-0.2.0 → autostatlib-0.2.2/src/AutoStatLib.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: AutoStatLib
-Version: 0.2.0
+Version: 0.2.2
 Summary: AutoStatLib - a simple statistical analysis tool
 Author: Stemonitis, SciWare LLC
 Author-email: konung-yaropolk <yaropolk1995@gmail.com>
@@ -531,6 +531,7 @@ License-File: LICENSE
 Requires-Dist: numpy
 Requires-Dist: scipy
 Requires-Dist: statsmodels
+Requires-Dist: pandas
 
 # AutoStatLib - python library for automated statistical analysis 
 
@@ -569,7 +570,7 @@ data_uniform = [list(np.random.uniform(i+3, i+1, n)) for i in range(groups)]
 
 
 # set the parameters:
-paired = False     # is groups dependend or not
+paired = False     # is groups dependent or not
 tails = 2          # two-tailed or one-tailed result
 popmean = 0        # population mean - only for single-sample tests needed
 
@@ -585,7 +586,7 @@ analysis.RunAuto()
 
 or you can choose specific tests:
 ```python
-# 2 groups independend:
+# 2 groups independent:
 analysis.RunTtest()
 analysis.RunMannWhitney()
 
@@ -594,10 +595,11 @@ analysis.RunTtestPaired()
 analysis.RunWilcoxon()
 
 # 3 and more independed groups comparison:
-analysis.RunAnova()
+analysis.RunOnewayAnova()
 analysis.RunKruskalWallis()
 
 # 3 and more depended groups comparison:
+analysis.RunOnewayAnovaRM()
 analysis.RunFriedman()
 
 # single group tests"
@@ -647,20 +649,40 @@ If errors occured, *GetResult()* returns an empty dictionary
 
 
 ---
-## Pre-Alpha dev status.
-
-### TODO:
-
---Kruskal-Wallis test - add Dunn's multiple comparisons
---Anova: add 2-way anova and 3-way(?)
-
-check:
---Wilcoxon signed-rank test and Mann-whitney - check mechanism of one-tailed calc, looks like it works wrong
-
-
-checked tests:
---Wilcoxon 2 tail - ok
---Mann-whitney 2 tail - ok
+## Pre-Alpha dev status.  
+  
+### TODO:  
+  
+-- Kruskal-Wallis test - add Dunn's multiple comparisons  
+-- Anova: add 2-way anova and 3-way anova  
+-- onevay Anova: add repeated measures (for normal dependent values) with and without Gaisser-Greenhouse correction  
+-- onevay Anova: add Brown-Forsithe and Welch (for normal independent values with unequal SDs between groups)  
+-- paired T-test: add ratio-paired t-test (ratios of paired values are consistent)  
+-- add Welch test (for norm data unequal variances)
+-- add Kolmogorov-smirnov test (unpaired nonparametric 2 sample, compare cumulative distributions)  
+-- add independent t-test with Welch correction (do not assume equal SDs in groups)  
+-- add correlation test, correlation diagram  
+-- add linear regression, regression diagram  
+-- add QQ plot
+-- n-sample tests: add onetail option
+  
+✅ done -- detailed normality test results
+
+  
+checked tests:  
+1-sample:  
+--Wilcoxon 2,1 tails - ok  
+--t-tests 2,1 tails -ok  
+  
+2-sample:  
+--Wilcoxon 2,1 tails - ok  
+--Mann-whitney 2,1 tails - ok  
+--t-tests 2,1 tails -ok  
+  
+n-sample:  
+--Kruskal-Wallis 2 tail - ok  
+--Friedman 2 tail - ok  
+--one-way ANOWA 2 tail - ok  
 
 
 

autostatlib-0.2.2/src/AutoStatLib.egg-info/requires.txt

@@ -0,0 +1,4 @@
+numpy
+scipy
+statsmodels
+pandas

autostatlib-0.2.0/requirements.txt

@@ -1,3 +0,0 @@
-numpy
-scipy
-statsmodels