upgrade pandas version

Author: AndreRicoPSU

clarite/internal/utilities.py

@@ -6,6 +6,9 @@
 import pandas as pd
 from pandas_genomics import GenotypeDtype
 
+# GITHUB ISSUE #120: SettingWithCopyWarning on Regression runs
+pd.set_option("mode.chained_assignment", None)
+
 
 def print_wrap(func):
     @wraps(func)
@@ -209,9 +212,8 @@ def _remove_empty_categories(
             if data[var].cat.ordered:
                 print()
             # GITHUB ISSUE #120: SettingWithCopyWarning on Regression runs
-            # data[var] = data[var].cat.remove_unused_categories()
-            data.loc[:, var] = data[var].cat.remove_unused_categories()
-
+            data[var] = data[var].cat.remove_unused_categories()
+            # data.loc[:, var] = data[var].cat.remove_unused_categories()
             removed_categories = set(existing_cats) - set(data[var].cat.categories)
             if len(removed_categories) > 0:
                 removed_cats[var] = removed_categories

clarite/modules/analyze/regression/base.py

@@ -105,7 +105,7 @@ def _validate_regression_params(self, regression_variables):
 
         # Collect lists of regression variables
         types = _get_dtypes(self.data)
-        rv_types = {v: t for v, t in types.iteritems() if v in regression_variables}
+        rv_types = {v: t for v, t in types.items() if v in regression_variables}
         rv_count = 0
         for dtype in ["binary", "categorical", "continuous", "genotypes"]:
             self.regression_variables[dtype] = [

clarite/modules/analyze/regression/glm_regression.py

@@ -380,6 +380,9 @@ def run(self):
                     )
                 )
 
+            # TODO: Error on multiprocess after update to Python > 3.8
+            self.process_num = 1
+
             if self.process_num == 1:
                 run_result = [
                     self._run_rv(

clarite/modules/analyze/regression/interaction_regression.py

@@ -263,6 +263,9 @@ def run(self):
             )
         )
 
+        # TODO: Error on multiprocess after update to Python > 3.8
+        self.process_num = 1
+
         if self.process_num == 1:
             run_result = [
                 self._run_interaction(

clarite/modules/analyze/regression/weighted_glm_regression.py

@@ -301,6 +301,9 @@ def run(self):
                     )
                 )
 
+            # TODO: Error on multiprocess after update to Python > 3.8
+            self.process_num = 1
+
             if self.process_num == 1:
                 run_result = [
                     self._run_weighted_rv(

clarite/modules/survey/survey_model.py

@@ -111,7 +111,7 @@ def _stata_linearization_vcov(self, X: pd.DataFrame, y: pd.DataFrame):
         ).set_index("clust", append=True)
 
         # Get sum of d_hat within each cluster (rows = clusters, columns = variables)
-        jdata = d_hat.groupby(axis=0, level="clust").apply(sum)
+        jdata = d_hat.groupby(axis=0, level="clust", group_keys=False).apply(sum)
 
         if self.design.has_strata:
             # Add strata label to jdata (just updates the index labels)
@@ -136,7 +136,7 @@ def center_strata(data, single_cluster_setting, pop_mean):
                     # This results in a value of 0 for the strata when there is only 1 cluster
                     return data - data.mean()
 
-            jdata = jdata.groupby(axis=0, level="strat").apply(
+            jdata = jdata.groupby(axis=0, level="strat", group_keys=False).apply(
                 lambda g: center_strata(g, self.design.single_cluster, d_hat.mean())
             )
 

docs/source/release-history.rst

@@ -2,6 +2,11 @@
 Release History
 ===============
 
+v2.3.1 (2023-01-10)
+-------------------
+
+Bug fixes (consult issues) and updating packages to Python 3.10
+
 v2.2.0 (2021-07-19)
 -------------------
 

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "clarite"
-version = "2.3.1"
+version = "2.3.2"
 description = "CLeaning to Analysis: Reproducibility-based Interface for Traits and Exposures"
 authors = ["Andre Rico <[email protected]>"]
 license = "BSD-3-Clause"
@@ -16,9 +16,9 @@ classifiers=[
     ]
 
 [tool.poetry.dependencies]
-python = " >=3.8.0,<3.11.0"
+python = ">=3.8.0,<3.11.0"
 click = ">7"
-pandas = "~1.3"
+pandas = "^1.3"
 scipy = "^1.9"
 seaborn = ">0.9"
 statsmodels = "^0.13"