Minor updates for v0.10.13 (#154)

* + update version

* + minor updates to copyright year

* + included ecdf option in eda.plot_int_dist

* + improved plot_int_dist

* + minor refactoring in plot_int_dist
+ minor bugfix in convert_dtypes

Author: jonsedar

LICENSE.md

@@ -3,7 +3,7 @@
 This License pertains to `oreum_core` hosted on
 [Oreum Industries' GitHub](https://github.com/oreum-industries/oreum_core)
 
-Copyright 2024 Oreum FZCO t/a Oreum Industries. All rights reserved.
+Copyright 2025 Oreum FZCO t/a Oreum Industries. All rights reserved.
 Oreum FZCO, IFZA, Dubai Silicon Oasis, Dubai, UAE, reg. 25515
 [oreum.io](https://oreum.io)
 

README.md

@@ -9,7 +9,6 @@
 [![code style: ruff](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/ruff/main/assets/badge/v2.json)](https://github.com/astral-sh/ruff)
 [![code style: interrogate](https://raw.githubusercontent.com/oreum-industries/oreum_core/master/assets/img/interrogate_badge.svg)](https://pypi.org/project/interrogate/)
 [![code security: bandit](https://img.shields.io/badge/code%20security-bandit-yellow.svg)](https://github.com/PyCQA/bandit)
-<!-- [![Conda Forge](https://img.shields.io/conda/vn/oreum-industries/oreum_core.svg)](https://anaconda.org/oreum-industries/oreum_core) -->
 
 ---
 

oreum_core/__init__.py

@@ -16,7 +16,7 @@
 
 import logging
 
-__version__ = "0.10.12"
+__version__ = "0.10.13"
 
 # logger goes to null handler by default
 # packages that import oreum_core can override this and direct elsewhere

oreum_core/curate/data_transform.py

@@ -113,8 +113,8 @@ def convert_dtypes(self, dfraw: pd.DataFrame) -> pd.DataFrame:
 
         for ft in self.ftsd["fbool"] + self.ftsd["fbool_nan_to_false"]:
             # tame string, strip, lower, use self.bool_dict, use pd.NA
-            # if not isinstance(df.dtypes[ft], bool):
-            if df.dtypes[ft].type != np.bool:
+            if not isinstance(df.dtypes[ft], bool):
+                # if df.dtypes[ft].type != bool:
                 df[ft] = df[ft].apply(lambda x: str(x).strip().lower())
                 df.loc[df[ft].isin(self.strnans), ft] = pd.NA
                 df[ft] = df[ft].apply(lambda x: self.bool_dict.get(x, x))
@@ -123,7 +123,7 @@ def convert_dtypes(self, dfraw: pd.DataFrame) -> pd.DataFrame:
                     df.loc[df[ft].isnull(), ft] = False
 
                 set_tf_only = set(df[ft].unique())
-                if set_tf_only in set([True, False]):  # most common, use np.bool
+                if len(set_tf_only - set([True, False])) == 0:  # most common
                     df[ft] = df[ft].astype(bool)
                 elif pd.isnull(df[ft]).sum() > 0:  # contains NaNs, use pd.boolean
                     df[ft] = df[ft].convert_dtypes(convert_boolean=True)

oreum_core/eda/plot.py

@@ -273,10 +273,19 @@ def plot_int_dist(
     vsize: float = 1.5,
     bins: int = None,
     plot_zeros: bool = True,
+    ecdf: bool = False,
     **kwargs,
 ) -> figure.Figure:
     """Plot group counts as histogram (optional log)"""
-    # handle under/over selecting fts
+    kws_hist = dict(stat="count")
+    legpos = "upper right"
+    t = "Empirical distribution"
+    if ecdf:
+        kws_hist = dict(stat="proportion", cumulative=True)
+        legpos = "lower right"
+        t += " ECDF"
+
+    # handles under/over selecting fts
     fts = list(set.intersection(set(df.columns.tolist()), set(fts)))
     if len(fts) == 0:
         return None
@@ -295,17 +304,16 @@ def plot_int_dist(
         ax = sns.histplot(
             df.loc[df[ft].notnull(), ft],
             kde=False,
-            stat="count",
             bins=bins,
             label=f"NaNs: {n_nans}, zeros: {n_zeros}, mean: {mean:.2f}, med: {med:.2f}",
             color=sns.color_palette()[i % 7],
             ax=ax1d[i][0],
+            **kws_hist,
         )
         if log:
-            _ = ax.set(yscale="log", title=ft, ylabel="log(count)")
-        _ = ax.set(title=ft, ylabel="count", xlabel=None)  # 'value'
-        _ = ax.legend(loc="upper right")
-    t = "Empirical distribution"
+            _ = ax.set(yscale="log", title=ft, ylabel=f"log({kws_hist['stat']})")
+        _ = ax.set(title=ft, ylabel=kws_hist["stat"], xlabel=None)
+        _ = ax.legend(loc=legpos)
     txtadd = kwargs.pop("txtadd", None)
     _ = f.suptitle(" - ".join(filter(None, [t, "ints", txtadd])), y=1, fontsize=14)
     _ = f.tight_layout(pad=0.9)