diffindiff 2.5.7__tar.gz → 2.5.8__tar.gz

{diffindiff-2.5.7 → diffindiff-2.5.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: diffindiff
-Version: 2.5.7
+Version: 2.5.8
 Summary: diffindiff: Python library for convenient Difference-in-Differences analyses
 Author: Thomas Wieland
 Author-email: geowieland@googlemail.com
@@ -40,14 +40,14 @@ Thomas Wieland [ORCID](https://orcid.org/0000-0001-5168-9846) [EMail](mailto:geo
 - 💻 GitHub Repository: [diffindiff_official](https://github.com/geowieland/diffindiff_official)
 - 📄 DOI (Zenodo): [10.5281/zenodo.18656820](https://doi.org/10.5281/zenodo.18656820)
 
-A case study that utilizes the diffindiff library is available on [arXiv](https://arxiv.org/abs/2605.21464).
+A case study that utilizes the diffindiff library (preprint) is available on [arXiv](https://arxiv.org/abs/2605.21464).
 
 
 ## Citation
 
 If you use this software, please cite:
 
-Wieland, T. (2026). diffindiff: A Python library for convenient difference-in-differences analyses (Version 2.5.7) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.18656820
+Wieland, T. (2026). diffindiff: A Python library for convenient difference-in-differences analyses (Version 2.5.8) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.18656820
 
 
 ## Installation
@@ -198,9 +198,12 @@ See the /tests directory for usage examples of most of the included functions.
 This software was developed without the use of AI-generated code. The GitHub Copilot Chat in Microsoft Visual Studio Code using the GPT-5 mini model (by OpenAI) was used solely to assist in drafting and refining docstrings for documentation. The corresponding guidelines and constraints defined by the author are documented in `AGENTS-docstrings.md` in the [public GitHub repository](https://github.com/geowieland/diffindiff_official).
 
 
-## What's new (v2.5.7)
+## What's new (v2.5.8)
 
 - Bugfixes
-  - Catching RecursionError when model formulas include too many variables
-  - Checking input data frames for duplicated columns
-  - Avoiding treatment group deviation in summary when treatment/control groups are identical
+  - didtools.is_balanced() now conducts a truly comprehensive check whether the input panel data is balanced
+  - Fixed bug in didtools.is_numeric() with respect to unintentional modification of the dataset
+  - diddata.create_counterfactual() now catches error if no model was trained by didtools.model_wrapper()
+  - didanalysis.DiffModel.plot_treatment_effects() now treats expections correctly
+- Other:
+  - verbose option in any function which prints NOTEs, with all NOTEs being disabled when verbose = False

{diffindiff-2.5.7 → diffindiff-2.5.8}/README.md

@@ -14,14 +14,14 @@ Thomas Wieland [ORCID](https://orcid.org/0000-0001-5168-9846) [EMail](mailto:geo
 - 💻 GitHub Repository: [diffindiff_official](https://github.com/geowieland/diffindiff_official)
 - 📄 DOI (Zenodo): [10.5281/zenodo.18656820](https://doi.org/10.5281/zenodo.18656820)
 
-A case study that utilizes the diffindiff library is available on [arXiv](https://arxiv.org/abs/2605.21464).
+A case study that utilizes the diffindiff library (preprint) is available on [arXiv](https://arxiv.org/abs/2605.21464).
 
 
 ## Citation
 
 If you use this software, please cite:
 
-Wieland, T. (2026). diffindiff: A Python library for convenient difference-in-differences analyses (Version 2.5.7) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.18656820
+Wieland, T. (2026). diffindiff: A Python library for convenient difference-in-differences analyses (Version 2.5.8) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.18656820
 
 
 ## Installation
@@ -172,9 +172,12 @@ See the /tests directory for usage examples of most of the included functions.
 This software was developed without the use of AI-generated code. The GitHub Copilot Chat in Microsoft Visual Studio Code using the GPT-5 mini model (by OpenAI) was used solely to assist in drafting and refining docstrings for documentation. The corresponding guidelines and constraints defined by the author are documented in `AGENTS-docstrings.md` in the [public GitHub repository](https://github.com/geowieland/diffindiff_official).
 
 
-## What's new (v2.5.7)
+## What's new (v2.5.8)
 
 - Bugfixes
-  - Catching RecursionError when model formulas include too many variables
-  - Checking input data frames for duplicated columns
-  - Avoiding treatment group deviation in summary when treatment/control groups are identical
+  - didtools.is_balanced() now conducts a truly comprehensive check whether the input panel data is balanced
+  - Fixed bug in didtools.is_numeric() with respect to unintentional modification of the dataset
+  - diddata.create_counterfactual() now catches error if no model was trained by didtools.model_wrapper()
+  - didanalysis.DiffModel.plot_treatment_effects() now treats expections correctly
+- Other:
+  - verbose option in any function which prints NOTEs, with all NOTEs being disabled when verbose = False

{diffindiff-2.5.7 → diffindiff-2.5.8}/diffindiff/config.py

@@ -4,15 +4,15 @@
 # Author:      Thomas Wieland 
 #              ORCID: 0000-0001-5168-9846
 #              mail: geowieland@googlemail.com              
-# Version:     1.0.28
-# Last update: 2026-10-10 09:58
+# Version:     1.0.29
+# Last update: 2026-10-17 21:17
 # Copyright (c) 2025-2026 Thomas Wieland
 #-----------------------------------------------------------------------
 
 # Basic config:
 
 PACKAGE_NAME = "diffindiff"
-PACKAGE_VERSION = "2.5.7"
+PACKAGE_VERSION = "2.5.8"
 
 VERBOSE = False
 

{diffindiff-2.5.7 → diffindiff-2.5.8}/diffindiff/didanalysis.py

@@ -4,8 +4,8 @@
 # Author:      Thomas Wieland 
 #              ORCID: 0000-0001-5168-9846
 #              mail: geowieland@googlemail.com              
-# Version:     2.4.6
-# Last update: 2026-09-11 21:46
+# Version:     2.4.8
+# Last update: 2026-10-17 21:15
 # Copyright (c) 2024-2026 Thomas Wieland
 #-----------------------------------------------------------------------
 
@@ -111,7 +111,8 @@ class DiffModel:
     def treatment_statistics(
         self,
         treatment: str = None,
-        after_treatment_col: str = None
+        after_treatment_col: str = None,
+        verbose: bool = False
         ):
 
         """
@@ -163,7 +164,9 @@ class DiffModel:
         else:
             
             treatment = model_config["treatment_col"][0]
-            print(f"NOTE: No treatment was stated. Choosing treatment '{treatment}' for calculating treatment statistics.")
+
+            if verbose:
+                print(f"NOTE: No treatment was stated. Choosing treatment '{treatment}' for calculating treatment statistics.")
 
         if after_treatment_col is not None:
             
@@ -782,9 +785,9 @@ class DiffModel:
       
         if len(no_control_conditions) > 0:
             if len(no_control_conditions) == 1:
-                print(f"NOTE: Treatment '{no_control_conditions[0]}' has no control conditions.")
+                print(f"WARNING: Treatment '{no_control_conditions[0]}' has no control conditions.")
             else:
-                print(f"NOTE: Treatments {', '.join(no_control_conditions)} have no control conditions.")  
+                print(f"WARNING: Treatments {', '.join(no_control_conditions)} have no control conditions.")  
 
         print("-" * int(total_width))        
 
@@ -905,6 +908,7 @@ class DiffModel:
         treatment_effects = self.treatment_effects()
 
         try:
+
             if model_config.get("pre_post", False) and len(model_config.get("after_treatment_col", [])) > 0:
 
                 labels = treatment_effects.iloc[:, 0].astype(str).values
@@ -943,9 +947,11 @@ class DiffModel:
                                     break
                         if appended:
                             break
-        except Exception:
+        except Exception as e:
 
-            pass
+            print(f"Plot of treatment effects not possible: {str(e)}")
+
+            return self
 
         if sort_by_coef:
             treatment_effects = treatment_effects.sort_values(
@@ -1099,7 +1105,8 @@ class DiffModel:
         self,
         treatment = None,
         after_treatment_col: str = None,
-        retransform_log_outcome: bool = False
+        retransform_log_outcome: bool = False,
+        verbose: bool = False
         ):
         
         """
@@ -1153,8 +1160,11 @@ class DiffModel:
             if treatment not in model_config["treatment_col"]:
                 raise ValueError (f"Treatment '{treatment}' not in model object.")
         else:
+            
             treatment = model_config["treatment_col"][0]
-            print(f"NOTE: No treatment was stated. Choosing treatment '{treatment}' for calculating counterfactual.")
+            
+            if verbose:
+                print(f"NOTE: No treatment was stated. Choosing treatment '{treatment}' for calculating counterfactual.")
 
         if after_treatment_col is not None:
             if len(model_config["after_treatment_col"]) == 0:
@@ -1313,7 +1323,8 @@ class DiffModel:
         TT_col: str = None,
         divide: float = 0.5,
         resample: float = 1.0,
-        random_state = 71
+        random_state = 71,
+        verbose: bool = False
         ):
 
         """
@@ -1369,8 +1380,11 @@ class DiffModel:
             if treatment not in model_config["treatment_col"]:
                 raise ValueError (f"Treatment '{treatment}' not in model object.")
         else:
+            
             treatment = model_config["treatment_col"][0]
-            print(f"NOTE: No treatment was stated. Choosing treatment '{treatment}' for placebo analysis.")
+            
+            if verbose:
+                print(f"NOTE: No treatment was stated. Choosing treatment '{treatment}' for placebo analysis.")
 
         if after_treatment_col is not None:
             if len(model_config["after_treatment_col"]) == 0:
@@ -1390,7 +1404,8 @@ class DiffModel:
             treatment_demeaned = treatment
             treatment = treatment.replace(f"{config.DELIMITER}{config.DEMEAN_SUFFIX}", "")
 
-            print(f"NOTE: Data is demeaned. Treatment statistics are calculated based on original treatment '{treatment}'.")
+            if verbose:
+                print(f"NOTE: Data is demeaned. Treatment statistics are calculated based on original treatment '{treatment}'.")
        
         treatment_statistics = self.treatment_statistics(treatment = treatment)
 
@@ -1484,7 +1499,8 @@ class DiffModel:
         plot_symbol: str = "o",
         treatment_group_only: bool = True,
         save_fig: str = None,
-        save_fig_kwargs: dict = {}
+        save_fig_kwargs: dict = {},
+        verbose: bool = False
         ):
 
         """
@@ -1542,8 +1558,11 @@ class DiffModel:
             if treatment not in model_config["treatment_col"]:
                 raise ValueError (f"Treatment '{treatment}' not in model object.")
         else:
+            
             treatment = model_config["treatment_col"][0]
-            print(f"NOTE: No treatment was stated. Choosing treatment '{treatment}' for plotting timeline.")
+            
+            if verbose:
+                print(f"NOTE: No treatment was stated. Choosing treatment '{treatment}' for plotting timeline.")
                 
         if treatment_group_only:
             if TG_col is None:
@@ -1615,7 +1634,8 @@ class DiffModel:
         pre_post_bar_width = 0.5,
         retransform_log_outcome: bool = False,
         save_fig: str = None,
-        save_fig_kwargs: dict = {}
+        save_fig_kwargs: dict = {},
+        verbose: bool = False
         ):
 
         """
@@ -1723,7 +1743,8 @@ class DiffModel:
                     break                        
         else:
             
-            print(f"NOTE: No treatment was stated. Choosing treatment '{treatment_diagnostics[0]['treatment']}' for plotting.")
+            if verbose:
+                print(f"NOTE: No treatment was stated. Choosing treatment '{treatment_diagnostics[0]['treatment']}' for plotting.")
 
             treatment_group = treatment_diagnostics[0]["treatment_group"]            
             treatment = treatment_diagnostics[0]["treatment"]
@@ -1783,13 +1804,15 @@ class DiffModel:
                 model_data[config.PREDICTIONS_SUMMARY_FRAME_COLS_LIST[4]] = np.exp(model_data[config.PREDICTIONS_SUMMARY_FRAME_COLS_LIST[4]])
                 model_data[config.PREDICTIONS_SUMMARY_FRAME_COLS_LIST[5]] = np.exp(model_data[config.PREDICTIONS_SUMMARY_FRAME_COLS_LIST[5]])
                 
-                if model_config["demean"]:
-                    print(f"NOTE: Outcome variable '{outcome_col.replace(f'{config.DELIMITER}{config.DEMEAN_SUFFIX}', '')}' was log-transformed and demeaned. Re-transformation was applied on '{outcome_col}'.")
-                else:
-                    print(f"NOTE: Outcome variable '{outcome_col}' was log-transformed. Re-transformation was applied.")
+                if verbose:
+                    if model_config["demean"]:
+                        print(f"NOTE: Outcome variable '{outcome_col.replace(f'{config.DELIMITER}{config.DEMEAN_SUFFIX}', '')}' was log-transformed and demeaned. Re-transformation was applied on '{outcome_col}'.")
+                    else:
+                        print(f"NOTE: Outcome variable '{outcome_col}' was log-transformed. Re-transformation was applied.")
             
             else:
-                print("NOTE: Parameter 'retransform_log_outcome' was set to True, but outcome variable was not log-transformed. No re-transformation applied.")
+                if verbose:
+                    print("NOTE: Parameter 'retransform_log_outcome' was set to True, but outcome variable was not log-transformed. No re-transformation applied.")
         
         model_data_TG = model_data[model_data[TG_col] == 1]
         model_data_CG = model_data[model_data[TG_col] == 0]
@@ -2096,7 +2119,8 @@ class DiffModel:
         plot_size: list = [12, 6],
         retransform_log_outcome: bool = False,
         save_fig: str = None,
-        save_fig_kwargs: dict = {}
+        save_fig_kwargs: dict = {},
+        verbose: bool = False
         ):
 
         """
@@ -2199,7 +2223,9 @@ class DiffModel:
                     break
     
         else:
-            print(f"NOTE: No treatment was stated. Choosing treatment '{treatment_diagnostics[0]['treatment']}' for plotting counterfactual.")
+
+            if verbose:
+                print(f"NOTE: No treatment was stated. Choosing treatment '{treatment_diagnostics[0]['treatment']}' for plotting counterfactual.")
 
             treatment_group = treatment_diagnostics[0]["treatment_group"]
 
@@ -2527,8 +2553,11 @@ def did_analysis(
         TG_col = []        
         
         if not FE_unit and config.AUTO_SWITCH_TO_FIXED_EFFECTS:
+
             FE_unit = True
-            print(f"NOTE: Quasi-experiment includes more than one treatment ({no_treatments}). Unit fixed effects are used instead of control group baseline and treatment group deviation.")
+            
+            if verbose:
+                print(f"NOTE: Quasi-experiment includes more than one treatment ({no_treatments}). Unit fixed effects are used instead of control group baseline and treatment group deviation.")
     
     if unique_units < config.FIXED_EFFECTS_THRESHOLD:
         
@@ -2538,7 +2567,8 @@ def did_analysis(
         ITT = False
         intercept = True
         
-        print(f"NOTE: Quasi-experiment includes less than {config.FIXED_EFFECTS_THRESHOLD} units. Unit or group fixed effects are skipped.")
+        if verbose:
+            print(f"NOTE: Quasi-experiment includes less than {config.FIXED_EFFECTS_THRESHOLD} units. Unit or group fixed effects are skipped.")
         
     if unique_time_points < config.FIXED_EFFECTS_THRESHOLD:
         
@@ -2547,7 +2577,8 @@ def did_analysis(
         GTT = False
         intercept = True
         
-        print(f"NOTE: Quasi-experiment includes less than {config.FIXED_EFFECTS_THRESHOLD} time points. Time fixed effects are skipped.")
+        if verbose:
+            print(f"NOTE: Quasi-experiment includes less than {config.FIXED_EFFECTS_THRESHOLD} time points. Time fixed effects are skipped.")
     
     if demean and FE_unit:
         
@@ -2556,38 +2587,56 @@ def did_analysis(
             ITE = False
             ITT = False
             
-            print(f"NOTE: Unit fixed effects are specified to be replaced by demeaning. {config.EFFECTS_TYPES['ITE']['description']} or {config.EFFECTS_TYPES['ITT']['description']} are not supported.")
+            if verbose:
+                print(f"NOTE: Unit fixed effects are specified to be replaced by demeaning. {config.EFFECTS_TYPES['ITE']['description']} or {config.EFFECTS_TYPES['ITT']['description']} are not supported.")
                        
     if ITE:
         
         if not FE_unit:
+
             FE_unit = True
-            print("NOTE: Model includes individual treatment effects. Unit fixed effects are included.")
+            
+            if verbose:
+                print("NOTE: Model includes individual treatment effects. Unit fixed effects are included.")
         
         if GTE:
+
             GTE = False
-            print("NOTE: Both group and individual treatment effects were stated. Switching to individual treatment effects only.")
+            
+            if verbose:
+                print("NOTE: Both group and individual treatment effects were stated. Switching to individual treatment effects only.")
     
     if ITT:
         
         TT_col = []
         
         if not FE_unit:
+
             FE_unit = True
-            print("NOTE: Model includes individual time trends. Unit fixed effects are included. Treatment time variable is dropped.")
+            
+            if verbose:
+                print("NOTE: Model includes individual time trends. Unit fixed effects are included. Treatment time variable is dropped.")
         
         if FE_time:
+            
             FE_time = False
-            print("NOTE: Model includes individual time trends. Time fixed effects are dropped.")
+            
+            if verbose:
+                print("NOTE: Model includes individual time trends. Time fixed effects are dropped.")
         
         if GTT:
+
             GTT = False
-            print("NOTE: Both group and individual time trends were stated. Switching to individual time trends only.")
+            
+            if verbose:
+                print("NOTE: Both group and individual time trends were stated. Switching to individual time trends only.")
             
     if staggered_adoption and config.AUTO_SWITCH_TO_FIXED_EFFECTS:
         
         if not FE_unit or not FE_time:
-            print("NOTE: Quasi-experiment includes one or more staggered treatments. Two-way fixed effects model is used.")
+
+            if verbose:
+                print("NOTE: Quasi-experiment includes one or more staggered treatments. Two-way fixed effects model is used.")
             
         FE_unit = True
         FE_time = True      
@@ -2604,7 +2653,9 @@ def did_analysis(
     if FE_group:
         
         if intercept or len(TG_col) > 0:
-            print("NOTE: Quasi-experiment includes group fixed effects. Control group baseline and treatment group deviation are dropped.")   
+
+            if verbose:
+                print("NOTE: Quasi-experiment includes group fixed effects. Control group baseline and treatment group deviation are dropped.")   
 
         TG_col = []
         intercept = False
@@ -2716,13 +2767,14 @@ def did_analysis(
         
         if config.AUTO_SKIP_CONSTANT_COLUMNS:
         
-            print(f"NOTE: The following columns are constant and dropped from the analysis: {', '.join(data_diagnostics['cols_constants'])}.")
+            print(f"WARNING: The following columns are constant and dropped from the analysis: {', '.join(data_diagnostics['cols_constants'])}.")
             
             covariates = [col for col in covariates if col not in data_diagnostics["cols_constants"]]
 
     if data_diagnostics["is_prepost"] and config.AUTO_SWITCH_TO_PREPOST:
         
-        print(f"NOTE: Input is {config.PREPOST_PANELDATA_DESCRIPTION}. Data processing and model estimation will treat data as pre-post.")
+        if verbose:
+            print(f"NOTE: Input is {config.PREPOST_PANELDATA_DESCRIPTION}. Data processing and model estimation will treat data as pre-post.")
         
         pre_post = True
         
@@ -2730,8 +2782,8 @@ def did_analysis(
         FE_time = False
         FE_group = False
     
-    if not data_diagnostics["is_prepost"] and not FE_unit and not FE_time:
-        print(f"NOTE: Input is {config.MULTIPERIOD_PANELDATA_DESCRIPTION}. Consider including {config.EFFECTS_TYPES['FE']['description']}.")
+    if not data_diagnostics["is_prepost"] and not FE_unit and not FE_time:        
+        print(f"WARNING: Input is {config.MULTIPERIOD_PANELDATA_DESCRIPTION}. Consider including {config.EFFECTS_TYPES['FE']['description']}.")
         
     if log_outcome:
 
@@ -2999,9 +3051,8 @@ def did_analysis(
         indep_vars_no = indep_vars_no-(len(FE_unit_vars)+len(FE_time_vars))
             
         if verbose:
-            print("OK")
-        
-        print(f"NOTE: Outcome variable '{outcome_col}' is replaced by demeaned variable '{outcome_col_demeaned}', and {len(treatment_col)+len(covariates)+len(spillover_vars)+len(interactions)} other variables were remeaned.")
+            print("OK")        
+            print(f"NOTE: Outcome variable '{outcome_col}' is replaced by demeaned variable '{outcome_col_demeaned}', and {len(treatment_col)+len(covariates)+len(spillover_vars)+len(interactions)} other variables were remeaned.")
         
         outcome_col = outcome_col_demeaned    
         
@@ -3283,13 +3334,15 @@ def ddd_analysis(
         
         FE_unit = False
           
-        print(f"NOTE: Quasi-experiment includes less than {config.FIXED_EFFECTS_THRESHOLD} units. Unit fixed effects are skipped.")
+        if verbose:
+            print(f"NOTE: Quasi-experiment includes less than {config.FIXED_EFFECTS_THRESHOLD} units. Unit fixed effects are skipped.")
         
     if unique_time_points < config.FIXED_EFFECTS_THRESHOLD:
         
         FE_time = False
         
-        print(f"NOTE: Quasi-experiment includes less than {config.FIXED_EFFECTS_THRESHOLD} time points. Time fixed effects are skipped.")
+        if verbose:
+            print(f"NOTE: Quasi-experiment includes less than {config.FIXED_EFFECTS_THRESHOLD} time points. Time fixed effects are skipped.")
                 
     if FE_unit:
         TG_col = None
@@ -3320,13 +3373,14 @@ def ddd_analysis(
         
         if config.AUTO_SKIP_CONSTANT_COLUMNS:
             
-            print(f"NOTE: The following columns are constant and dropped from the analysis: {', '.join(data_diagnostics['cols_constants'])}.")
+            print(f"WARNING: The following columns are constant and dropped from the analysis: {', '.join(data_diagnostics['cols_constants'])}.")
             
             covariates = [col for col in covariates if col not in data_diagnostics["cols_constants"]]
     
     if data_diagnostics["is_prepost"] and config.AUTO_SWITCH_TO_PREPOST:
         
-        print("NOTE: Panel data is pre-post. Data processing and model estimation will treat data as pre-post")
+        if verbose:
+            print("NOTE: Panel data is pre-post. Data processing and model estimation will treat data as pre-post")
         
         pre_post = True
         

{diffindiff-2.5.7 → diffindiff-2.5.8}/diffindiff/didanalysis_helper.py

@@ -4,8 +4,8 @@
 # Author:      Thomas Wieland 
 #              ORCID: 0000-0001-5168-9846
 #              mail: geowieland@googlemail.com              
-# Version:     1.2.5
-# Last update: 2026-10-09 23:04
+# Version:     1.2.6
+# Last update: 2026-10-17 18:55
 # Copyright (c) 2025-2026 Thomas Wieland
 #-----------------------------------------------------------------------
 
@@ -231,8 +231,8 @@ def demean_variables(
     suffix = f"{config.DELIMITER}{config.DEMEAN_SUFFIX}"
 
     if not FE_unit and not FE_time:
-        
-        print("NOTE: Neither 'FE_unit' nor 'FE_time' is set to True. No demeaning will be performed.")
+
+        print("WARNING: Neither 'FE_unit' nor 'FE_time' is set to True. No demeaning will be performed.")
         
         for col in cols_to_demean:
 
@@ -258,13 +258,13 @@ def demean_variables(
     else:
         
         tools.check_columns(
-            df = data,
+            df = data_demean,
             columns = cols_to_demean + [unit_col] + [time_col],
             verbose = verbose
             )
                 
         tools.is_numeric(
-            df = data,
+            df = data_demean,
             columns = cols_to_demean,
             verbose = verbose
             )

{diffindiff-2.5.7 → diffindiff-2.5.8}/diffindiff/diddata.py

@@ -4,8 +4,8 @@
 # Author:      Thomas Wieland 
 #              ORCID: 0000-0001-5168-9846
 #              mail: geowieland@googlemail.com              
-# Version:     2.3.3
-# Last update: 2026-08-04 17:51
+# Version:     2.3.4
+# Last update: 2026-10-17 21:34
 # Copyright (c) 2024-2026 Thomas Wieland
 #-----------------------------------------------------------------------
 
@@ -1763,7 +1763,7 @@ class DiffData:
         TT_col = treatment_cols[0]["TT_col"]
         treatment_name = treatment_cols[0]["treatment_name"]
 
-        if len(treatment_cols) > 1:
+        if len(treatment_cols) > 1 and verbose:
             print(f"NOTE: Treatment data contains {len(treatment_cols)} treatments. Taking first treatment '{treatment_name}' for identifying treatment group and time.")
         
         if verbose:
@@ -1853,9 +1853,8 @@ class DiffData:
         self.data[1] = groups
 
         if verbose:
-            print("OK")
-        
-        print("NOTE: DiffData object was changed with counterfactual.")
+            print("OK")        
+            print("NOTE: DiffData object was changed with counterfactual.")
 
         return self
 
@@ -1926,7 +1925,7 @@ class DiffData:
         treatment_name = treatment_cols[0]["treatment_name"]
         TG_col = treatment_cols[0]["TG_col"]
         
-        if len(treatment_cols) > 1:
+        if len(treatment_cols) > 1 and verbose:
             print(f"NOTE: Treatment data contains {len(treatment_cols)} treatments. Taking first treatment '{treatment_name}' for identifying treatment group and time.")
         
         if verbose:
@@ -2014,7 +2013,7 @@ class DiffData:
         units = valid_units
 
         if len(units_not_included) > 0:
-            print(f"WARNING: {len(units_not_included)} analysis units were not included due to NaN data.")
+            print(f"WARNING: {len(units_not_included)} analysis units were not included due to NaN values.")
 
         if verbose:
             print("OK")
@@ -2863,7 +2862,7 @@ def create_counterfactual(
     data_len = len(data)
     data = data.dropna()
     if len(data) < data_len:
-        print(f"NOTE: Because of NaN values, {data_len-len(data)} observations were skipped.")
+        print(f"WARNING: Because of NaN values, {data_len-len(data)} observations were skipped.")
 
     isnotreatment = tools.is_notreatment(
         data = data,
@@ -2871,7 +2870,7 @@ def create_counterfactual(
         treatment_col = treatment_col,
         verbose = verbose
         )
-    if not isnotreatment[0]:
+    if not isnotreatment[0] and verbose:
         print(f"NOTE: No {config.NO_TREATMENT_CG_DESCRIPTION}. Counterfactual will not cover full treatment time.")    
     
     data_unit = data.loc[data[unit_col].astype(str) == cf_for_unit]
@@ -2897,7 +2896,7 @@ def create_counterfactual(
     units = tools.unique(units_tt[unit_col])  
     
     if not isnotreatment[0]:
-        print(f"NOTE: No {config.TREATMENT_GROUP_DESCRIPTION} units in model data left.")
+        print(f"WARNING: No {config.TREATMENT_GROUP_DESCRIPTION} units in model data left.")
     
     if verbose:
         print(f"Creating data for counterfactual with {len(X)} independent variables", end = " ... ")
@@ -2958,18 +2957,30 @@ def create_counterfactual(
         random_state = random_state,
         verbose = verbose
         )
-    
-    if verbose:
-        print(f"Counterfactual prediction for observational unit '{cf_for_unit}'", end = " ... ")
-    
-    data_unit[treatment_col] = 0
-    data_unit[y] = counterfactual_pred[1].predict(data_unit[X])
-    
-    if verbose:
-        print("OK")
+
+    model_estimation_error = counterfactual_pred[6]
+    model_estimation_error_text = ""
+
+    if model_estimation_error:
+
+        model_estimation_error_text = counterfactual_pred[7]
+        print(f"WARNING: The desired {config.MODEL_WRAPPER_AVAILABLE[model_type]} model was not estimated. The following error occurred: '{model_estimation_error_text}'.")
+
+    else:
+
+        if verbose:
+            print(f"Counterfactual prediction for observational unit '{cf_for_unit}'", end = " ... ")
+        
+        data_unit[treatment_col] = 0
+        data_unit[y] = counterfactual_pred[1].predict(data_unit[X])
+        
+        if verbose:
+            print("OK")
     
     return [
         counterfactual_pred, 
         data_cf, 
-        data_unit
+        data_unit,
+        model_estimation_error,
+        model_estimation_error_text
         ]

{diffindiff-2.5.7 → diffindiff-2.5.8}/diffindiff/didtools.py

@@ -4,8 +4,8 @@
 # Author:      Thomas Wieland 
 #              ORCID: 0000-0001-5168-9846
 #              mail: geowieland@googlemail.com              
-# Version:     2.2.7
-# Last update: 2026-10-10 09:57
+# Version:     2.2.8
+# Last update: 2026-10-17 18:53
 # Copyright (c) 2025-2026 Thomas Wieland
 #-----------------------------------------------------------------------
 
@@ -77,6 +77,8 @@ def check_columns(
     """
 
     if len(columns) > 0:
+
+        key_errors = []
         
         if verbose:
             print(f"Checking if column(s) {', '.join(columns)} exist(s) in data frame", end=" ... ")
@@ -87,7 +89,7 @@ def check_columns(
             print("OK")
         
         if missing_columns:
-            raise KeyError(f"Data do not contain column(s): {', '.join(missing_columns)}")
+            key_errors.append(f"Data do not contain column(s): {', '.join(missing_columns)}")
 
         if verbose:
             print("Checking whether columns are duplicated in data frame", end = " ... ")
@@ -98,7 +100,10 @@ def check_columns(
             print("OK")
 
         if len(cols_duplicated) > 0:
-            raise KeyError(f"Data contain duplicated relevant columns: {', '.join(cols_duplicated)}")
+            key_errors.append(f"Data contain duplicated relevant columns: {', '.join(cols_duplicated)}")
+
+        if len(key_errors) > 0:
+            raise KeyError(f"{' '.join(key_errors)}")
 
 def is_numeric(
     df: pd.DataFrame, 
@@ -134,6 +139,8 @@ def is_numeric(
     >>> is_numeric(df, ['x'])
     """
 
+    df = df.copy()
+
     if len(columns) > 0:
         
         if verbose:
@@ -266,32 +273,41 @@ def is_balanced(
     if other_cols is None:
         other_cols = []
 
+    cols_relevant = [unit_col, time_col, outcome_col] + other_cols
+
     if verbose:
-        print(f"Checking whether panel data is balanced with respect to {3+len(other_cols)} columns", end = " ... ")
+        print(f"Checking whether panel data is balanced with respect to {len(cols_relevant)} columns", end = " ... ")
 
-    unit_freq = data[unit_col].nunique()
-    time_freq = data[time_col].nunique()
-    unitxtime = unit_freq*time_freq
+    panel = data[cols_relevant].copy()
 
-    if other_cols is None:
-        cols_relevant = [unit_col, time_col, outcome_col]
-    else:
-        cols_relevant = [unit_col, time_col, outcome_col] + other_cols
+    warnings = []
 
-    data_relevant = data[cols_relevant]
-    
-    if unitxtime != len(data_relevant.notna()):
-        balanced = False
-    else:
-        balanced = True
-    
+    if panel.isnull().any().any():
+        warnings.append("Relevant columns contain missing values.")
+        
+    pair_counts = panel[[unit_col, time_col]].dropna().groupby([unit_col, time_col]).size()
+    if (pair_counts != 1).any():
+        warnings.append("Unit-time pairs are duplicated or incomplete.")
+
+    units = panel[unit_col].dropna().unique()
+    times = panel[time_col].dropna().unique()
+    expected_pairs = set((u, t) for u in units for t in times)
+    observed_pairs = set(pair_counts.index.tolist())
+    if observed_pairs != expected_pairs:
+        warnings.append("Some unit-time combinations are missing.")
+        
     if verbose:
         print("OK")
 
-    if not balanced:
-        print("WARNING: Panel data is not balanced.")
-        
-    return balanced
+    if len(warnings) > 0:
+
+        if verbose:
+            print(f"WARNING: {' '.join(warnings)}")
+
+        return False
+    
+    else:
+        return True
 
 def is_dummy(
     iterable,
@@ -562,7 +578,7 @@ def is_simultaneous(
         if verbose:
             print("OK")
 
-        if not simultaneous and data_isnotreatment[0]:
+        if not simultaneous and data_isnotreatment[0] and verbose:
             print(f"NOTE: Treatment '{treatment_col}' is not simultaneous.")
 
         if simultaneous and not data_isnotreatment[0]:
@@ -621,8 +637,8 @@ def is_notreatment(
     if verbose:
         print("OK")
 
-        if not no_treatment:
-            print(f"NOTE: treatment '{treatment_col}' does not include a {config.NO_TREATMENT_CG_DESCRIPTION}.")
+    if not no_treatment:
+        print(f"WARNING: treatment '{treatment_col}' does not include a {config.NO_TREATMENT_CG_DESCRIPTION}.")
 
     return [
         no_treatment, 
@@ -1708,7 +1724,9 @@ def model_wrapper(
             X_test, 
             y_train, 
             y_test
-        )
+        ),
+        model_estimation_error,
+        model_estimation_error_text
         ]
 
 def fit_metrics(
@@ -1824,8 +1842,7 @@ def fit_metrics(
  
     if any(observed == 0):
         
-        if verbose:
-            print ("NOTE: Vector 'observed' contains values equal to zero. No APE/MAPE calculated.")
+        print ("WARNING: Vector 'observed' contains values equal to zero. No APE/MAPE calculated.")
         
         APE = np.full_like(observed, np.nan)
         MAPE = np.nan

{diffindiff-2.5.7 → diffindiff-2.5.8}/diffindiff/tests/tests_diffindiff.py

@@ -4,8 +4,8 @@
 # Author:      Thomas Wieland 
 #              ORCID: 0000-0001-5168-9846
 #              mail: geowieland@googlemail.com              
-# Version:     2.2.0
-# Last update: 2026-08-04 18:42
+# Version:     2.2.1
+# Last update: 2026-10-17 14:50
 # Copyright (c) 2025-2026 Thomas Wieland
 #-----------------------------------------------------------------------
 
@@ -305,7 +305,6 @@ curfew_model_FE.plot(
 
 fixed_effects = curfew_model_FE.fixed_effects()
 # Fixed effects of CHBW_data_model_FE
-#print(fixed_effects)
 
 
 # Model with demeaned variables:

{diffindiff-2.5.7 → diffindiff-2.5.8}/diffindiff.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: diffindiff
-Version: 2.5.7
+Version: 2.5.8
 Summary: diffindiff: Python library for convenient Difference-in-Differences analyses
 Author: Thomas Wieland
 Author-email: geowieland@googlemail.com
@@ -40,14 +40,14 @@ Thomas Wieland [ORCID](https://orcid.org/0000-0001-5168-9846) [EMail](mailto:geo
 - 💻 GitHub Repository: [diffindiff_official](https://github.com/geowieland/diffindiff_official)
 - 📄 DOI (Zenodo): [10.5281/zenodo.18656820](https://doi.org/10.5281/zenodo.18656820)
 
-A case study that utilizes the diffindiff library is available on [arXiv](https://arxiv.org/abs/2605.21464).
+A case study that utilizes the diffindiff library (preprint) is available on [arXiv](https://arxiv.org/abs/2605.21464).
 
 
 ## Citation
 
 If you use this software, please cite:
 
-Wieland, T. (2026). diffindiff: A Python library for convenient difference-in-differences analyses (Version 2.5.7) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.18656820
+Wieland, T. (2026). diffindiff: A Python library for convenient difference-in-differences analyses (Version 2.5.8) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.18656820
 
 
 ## Installation
@@ -198,9 +198,12 @@ See the /tests directory for usage examples of most of the included functions.
 This software was developed without the use of AI-generated code. The GitHub Copilot Chat in Microsoft Visual Studio Code using the GPT-5 mini model (by OpenAI) was used solely to assist in drafting and refining docstrings for documentation. The corresponding guidelines and constraints defined by the author are documented in `AGENTS-docstrings.md` in the [public GitHub repository](https://github.com/geowieland/diffindiff_official).
 
 
-## What's new (v2.5.7)
+## What's new (v2.5.8)
 
 - Bugfixes
-  - Catching RecursionError when model formulas include too many variables
-  - Checking input data frames for duplicated columns
-  - Avoiding treatment group deviation in summary when treatment/control groups are identical
+  - didtools.is_balanced() now conducts a truly comprehensive check whether the input panel data is balanced
+  - Fixed bug in didtools.is_numeric() with respect to unintentional modification of the dataset
+  - diddata.create_counterfactual() now catches error if no model was trained by didtools.model_wrapper()
+  - didanalysis.DiffModel.plot_treatment_effects() now treats expections correctly
+- Other:
+  - verbose option in any function which prints NOTEs, with all NOTEs being disabled when verbose = False

{diffindiff-2.5.7 → diffindiff-2.5.8}/diffindiff.egg-info/SOURCES.txt

@@ -16,4 +16,5 @@ diffindiff/tests/__init__.py
 diffindiff/tests/tests_diffindiff.py
 diffindiff/tests/data/Corona_Hesse.xlsx
 diffindiff/tests/data/counties_DE.csv
-diffindiff/tests/data/curfew_DE.csv
+diffindiff/tests/data/curfew_DE.csv
+tests/test_diffindiff.py

{diffindiff-2.5.7 → diffindiff-2.5.8}/setup.py

@@ -7,7 +7,7 @@ def read_README():
     
 setup(
     name='diffindiff',
-    version='2.5.7',
+    version='2.5.8',
     description='diffindiff: Python library for convenient Difference-in-Differences analyses',
     packages=find_packages(include=["diffindiff", "diffindiff.tests"]),
     include_package_data=True,

diffindiff-2.5.8/tests/test_diffindiff.py

@@ -0,0 +1,257 @@
+#-----------------------------------------------------------------------
+# Name:        tests (diffindiff package)
+# Purpose:     Tests for diffindiff package functions
+# Author:      Thomas Wieland 
+#              ORCID: 0000-0001-5168-9846
+#              mail: geowieland@googlemail.com              
+# Version:     1.0.0
+# Last update: 2026-10-17 15:01
+# Copyright (c) 2026 Thomas Wieland
+#-----------------------------------------------------------------------
+
+
+from pathlib import Path
+import pandas as pd
+
+
+def test_import_diffindiff():
+    """Test if diffindiff package can be imported."""
+
+    import diffindiff
+    from diffindiff.config import PACKAGE_NAME, PACKAGE_VERSION
+
+    print(f"This is the {PACKAGE_NAME} package v{PACKAGE_VERSION}")
+
+
+def test_prepost_analysis():
+
+    from diffindiff.diddata import create_groups, create_treatment, merge_data, create_data
+   
+    curfew_DE = pd.read_csv(
+        Path(__file__).parent / "data" / "curfew_DE.csv",
+        sep=";",
+        decimal=","
+    )
+    # Dataset with daily and cumulative SARS-CoV-2 infections of German counties
+    # Data source: Wieland (2020) https://doi.org/10.18335/region.v7i2.324
+
+    curfew_groups=create_groups(
+        treatment_group= 
+            curfew_DE.loc[curfew_DE["Bundesland"].isin([9,10,14])]["county"],
+        control_group= 
+            curfew_DE.loc[~curfew_DE["Bundesland"].isin([9,10,14])]["county"]
+        )
+    # Creating treatment and control group
+    # "Bundesland" (=federal state): 9 = Bavaria, 10 = Saarland, 14 = Saxony
+
+    assert curfew_groups is not None
+
+    curfew_groups.summary()
+    # Groups summary
+
+    curfew_treatment_prepost=create_treatment(
+        study_period=["2020-03-01", "2020-05-15"],
+        treatment_period=["2020-03-21", "2020-05-05"],
+        freq = "D",
+        pre_post = True
+        )
+    # Creating treatment
+    # Curfew from March 21, 2020, to May, 5, 2020
+
+    assert curfew_treatment_prepost is not None
+
+    curfew_treatment_prepost.summary()
+    # Treatment summary
+
+    curfew_data_prepost_merge=merge_data(
+        outcome_data=curfew_DE,
+        unit_id_col="county",
+        time_col="infection_date",
+        outcome_col="infections_cum_per100000",
+        diff_groups=curfew_groups,
+        diff_treatment=curfew_treatment_prepost
+        )
+
+    assert curfew_data_prepost_merge is not None
+
+    curfew_data_prepost_merge.summary()
+    # Summary of created data
+
+    curfew_data_prepost=create_data(
+        outcome_data=curfew_DE,
+        unit_id_col="county",
+        time_col="infection_date",
+        outcome_col="infections_cum_per100000",
+        treatment_group= 
+            curfew_DE.loc[curfew_DE["Bundesland"].isin([9,10,14])]["county"],
+        control_group= 
+            curfew_DE.loc[~curfew_DE["Bundesland"].isin([9,10,14])]["county"],
+        study_period=["2020-03-01", "2020-05-15"],
+        treatment_period=["2020-03-21", "2020-05-05"],
+        freq="D",
+        pre_post=True,
+        after_treatment_period = False
+        )
+    # Creating DiD treatement dataset by defining groups and
+    # treatment time at once
+
+    assert curfew_data_prepost is not None
+
+    curfew_data_prepost.summary()
+    # Summary of created data
+
+    curfew_model_prepost=curfew_data_prepost.analysis(verbose=True)
+    # Model analysis of created data
+
+    assert curfew_model_prepost is not None
+
+    print(curfew_model_prepost.treatment_effects())
+    # Show treatment effects
+
+    print(curfew_model_prepost.fixed_effects())
+    # Show fixed effects (not included in model)
+
+    print(curfew_model_prepost.covariates())
+    # Show covariates (not included in model)
+
+    curfew_model_prepost.summary()
+    # Model summary
+
+    print(curfew_model_prepost.didmodel().summary())
+    # Access statsmodels OLS object
+
+    print(curfew_model_prepost.fit_metrics())
+    # Show model fit metrics
+
+    curfew_model_prepost.plot(
+        x_label="Timepoint",
+        y_label="Cumulative infections per 100,000",
+        plot_title="Curfew effectiveness pre-post - Groups over time",
+        plot_observed=False,
+        lines_col=[None,None,"blue","orange"],
+        lines_labels=[None,None,"Treatment group","Control group","Treatment group CI","Control group CI"],
+        lines_style=[None,None,"solid","solid"]
+        )
+    # Plot DiD pre vs. post results
+    # with user-determined style
+
+    curfew_model_prepost.plot(
+        x_label="Timepoint",
+        y_label="Cumulative infections per 100,000",
+        plot_title="Curfew effectiveness pre-post - Groups over time",
+        lines_col=[None,None,"blue","orange"],
+        lines_labels=[None,None,"Treatment group","Control group","Treatment group CI", "Control group CI"],
+        pre_post_barplot=True
+        )
+    # Plot DiD pre vs. post results
+    # with user-determined style
+
+    curfew_model_prepost.plot_treatment_effects(
+        x_label="Coefficients with 95% CI",
+        plot_title="Curfew effectiveness pre-post - DiD effects"
+        )
+    # plot effects
+
+    curfew_model_prepost.plot_treatment_effects(
+        x_label="Coefficients with 95% CI",
+        plot_title="Curfew effectiveness pre-post - DiD effects",
+        scale_plot=False
+        )
+    # plot effects
+
+def test_multiperiod_paneldata_analysis():
+
+    from diffindiff.diddata import create_data
+       
+    curfew_DE = pd.read_csv(
+        Path(__file__).parent / "data" / "curfew_DE.csv",
+        sep=";",
+        decimal=","
+    )
+    # Dataset with daily and cumulative SARS-CoV-2 infections of German counties
+    # Data source: Wieland (2020) https://doi.org/10.18335/region.v7i2.324
+
+    curfew_data=create_data(
+        outcome_data=curfew_DE,
+        unit_id_col="county",
+        time_col="infection_date",
+        outcome_col="infections_cum_per100000",
+        treatment_group= 
+            curfew_DE.loc[curfew_DE["Bundesland"].isin([9,10,14])]["county"],
+        control_group= 
+            curfew_DE.loc[~curfew_DE["Bundesland"].isin([9,10,14])]["county"],
+        treatment_name="Curfew",
+        study_period=["2020-03-01", "2020-05-15"],
+        treatment_period=["2020-03-21", "2020-05-05"],
+        freq="D"
+        )
+    # Creating DiD dataset by defining groups and treatment time at once
+    # Treatment curfew in three federal states: Bavaria (9), Saarland (10) and Saxony (14)
+   
+    assert curfew_data is not None
+
+    curfew_data.summary()
+    # Summary of created treatment data
+
+    curfew_model=curfew_data.analysis(verbose=True)
+    # Model analysis of created data
+
+    assert curfew_model is not None
+
+    curfew_model.summary()
+    # Model summary
+
+    curfew_model.plot(
+        y_label="Cumulative infections per 100,000",
+        plot_title="Curfew effectiveness - Groups over time",
+        plot_observed=True
+        )
+    # Plot observed vs. predicted (means) separated by group (treatment and control)
+
+    curfew_model.plot_treatment_effects(
+        x_label="Coefficients with 95% CI",
+        plot_title="Curfew effectiveness - DiD effects"
+        )
+    # plot effects
+
+    curfew_placebo = curfew_model.placebo(
+        treatment="Curfew"
+        )
+    # Placebo test with default parameters
+
+    assert curfew_placebo is not None
+
+    curfew_placebo.summary()
+    # Summary of placebo test
+
+    # Two-way-fixed-effects model:
+
+    curfew_model_FE=curfew_data.analysis(
+        FE_unit=True, 
+        FE_time=True,
+        demean = True,
+        verbose = True
+        )
+    # Model analysis of created data with fixed effects for 
+    # units (FE_unit=True) and time (FE_time=True)
+    # data is demeaned
+
+    assert curfew_model_FE is not None
+
+    curfew_model_FE_summary = curfew_model_FE.summary()
+    # Model summary
+
+    assert curfew_model_FE_summary is not None
+
+    print(curfew_model_FE.treatment_statistics())
+    # Print treatment statistics
+
+    curfew_model_FE.plot(
+        y_label="Cumulative infections per 100,000",
+        plot_title="Curfew effectiveness (Two-way FE model) - Groups over time",
+        plot_observed=True
+        )
+    # Plot of treatment and control group
+
+    fixed_effects = curfew_model_FE.fixed_effects()
+    # Fixed effects of CHBW_data_model_FE