fixed major bugs in stairs, report and plot_heatmap methods

Author: maximtrp

setup.cfg

@@ -13,12 +13,12 @@ classifiers =
     License :: OSI Approved :: MIT License
     Operating System :: OS Independent
     Programming Language :: Python :: 3
-    Programming Language :: Python :: 3.5
     Programming Language :: Python :: 3.6
     Programming Language :: Python :: 3.7
     Programming Language :: Python :: 3.8
     Programming Language :: Python :: 3.9
     Programming Language :: Python :: 3.10
+    Programming Language :: Python :: 3.11
     Topic :: Scientific/Engineering :: Information Analysis
 
 [options]

src/scikit_na/__init__.py

@@ -4,4 +4,4 @@
 from ._stats import *  # noqa: F401, F403
 from ._report import *  # noqa: F401, F403
 
-__version__ = '0.1.0'
+__version__ = '0.1.1'

src/scikit_na/_report.py

@@ -114,7 +114,7 @@ def _on_vis_col_select(names):
 
     select_vis_cols = widgets.SelectMultiple(options=cols, rows=6)
     select_vis_cols.observe(_on_vis_col_select, names='value')
-    select_vis_accordion = widgets.Accordion(children=[select_cols])
+    select_vis_accordion = widgets.Accordion(children=[select_vis_cols])
     select_vis_accordion.set_title(0, 'Columns selection')
     select_vis_accordion.selected_index = 0
 
@@ -147,6 +147,7 @@ def _on_vis_col_select(names):
     def _on_stats_col_select(_):
         # Clearing display
         stats_table.clear_output(wait=False)
+        stats_table2.clear_output(wait=False)
         # Getting selected column with NA values
         _col_na = select_stats_col_na.value
         # Getting selected columns
@@ -157,15 +158,21 @@ def _on_stats_col_select(_):
         _cols_nominal = _get_nominal_cols(data, _cols)
 
         with stats_table:
-            display(
-                describe(
-                    data, col_na=_col_na, columns=_cols_numeric)
-                .round(round_dec))
+            try:
+                display(
+                    describe(
+                        data, col_na=_col_na, columns=_cols_numeric)
+                    .round(round_dec))
+            except:
+                display(widgets.HTML('Please select numeric columns to describe'))
         with stats_table2:
-            display(
-                describe(
-                    data, col_na=_col_na, columns=_cols_nominal)
-                .round(round_dec))
+            try:
+                display(
+                    describe(
+                        data, col_na=_col_na, columns=_cols_nominal)
+                    .round(round_dec))
+            except:
+                display(widgets.HTML('Please select nominal columns to describe'))
 
     # Setting up dropdown and select elements for choosing columns
     select_stats_col_na_header = widgets.HTML(

src/scikit_na/_stats.py

@@ -15,7 +15,7 @@ def _select_cols(
     return array(
         list(col for col in columns)
         if columns is not None
-        else (data.columns if second_var is None else second_var))
+        else (data.columns if second_var is None or len(second_var) == 0 else second_var))
 
 
 def _get_nominal_cols(
@@ -106,7 +106,7 @@ def summary(
     """
     cols = _select_cols(data, columns)
     data_copy = data.loc[:, cols].copy()
-    na_by_inst = (data_copy.isna().sum(axis=1) == 1)
+    na_by_inst = data_copy.isna().sum(axis=1) == 1
     na_total = _get_total_na_count(data_copy, cols)
 
     if per_column:
@@ -201,8 +201,8 @@ def stairs(
         Dataset shrinkage results after cumulative
         :py:meth:`pandas.DataFrame.dropna()`.
     """
-    data_copy = data.copy()
     cols = _select_cols(data, columns).tolist()
+    data_copy = data.loc[:, cols].copy()
     stairs_values = []
     stairs_labels = []
 

src/scikit_na/altair/_altair.py

@@ -5,7 +5,7 @@
 from typing import Optional, Sequence
 from numbers import Integral
 from ipywidgets import widgets, interact
-from numpy import arange, r_, nan, fill_diagonal
+from numpy import arange, nan, fill_diagonal
 from pandas import DataFrame
 from altair import (
     Axis, Chart, Color, condition, data_transformers, selection_multi,
@@ -401,7 +401,7 @@ def plot_stairs(
             x=X(**x_kws),
             y=Y(**y_kws),
             tooltip=[xlabel, ylabel, tooltip_label]
-        )
+    )
     return chart\
         .configure_axis(labelFontSize=font_size, titleFontSize=font_size)\
         .configure_legend(labelFontSize=font_size, titleFontSize=font_size)
@@ -472,17 +472,15 @@ def plot_stairbars(
             x=X(**x_kws),
             y=Y(**y_kws),
             tooltip=[xlabel, ylabel, tooltip_label]
-        )
+    )
     return chart\
         .configure_axis(labelFontSize=font_size, titleFontSize=font_size)\
         .configure_legend(labelFontSize=font_size, titleFontSize=font_size)
 
 
-
 def plot_heatmap(
         data: DataFrame,
         columns: Optional[Sequence[str]] = None,
-        tooltip_cols: Optional[Sequence[str]] = None,
         names: list = None,
         sort: bool = True,
         droppable: bool = True,
@@ -506,8 +504,6 @@ def plot_heatmap(
         Input data.
     columns : Optional[Sequence[str]], optional
         Columns that are to be displayed on a plot.
-    tooltip_cols : Optional[Sequence[str]], optional
-        Columns to be used in tooltips.
     names : list, optional
         Values labels passed as a list.
         The first element corresponds to non-missing values,
@@ -545,7 +541,8 @@ def plot_heatmap(
     if not x_kws:
         x_kws = {'sort': None, 'shorthand': xlabel, 'type': 'nominal'}
     if not y_kws:
-        y_kws = {'sort': None, 'shorthand': ylabel, 'type': 'ordinal', 'axis': Axis(labelOverlap='greedy')}
+        y_kws = {'sort': None, 'shorthand': ylabel,
+                 'type': 'ordinal', 'axis': Axis(labelOverlap='greedy')}
     if not names:
         names = ['Filled', 'NA', 'Droppable']
     if not color_kws:
@@ -557,37 +554,35 @@ def plot_heatmap(
                 range=["green", "red", "orange"])
         }
     if not rect_kws:
-        rect_kws = {}
+        rect_kws = {"clip": True}
 
     cols = _select_cols(data, columns)
-    tt_cols = _select_cols(data, tooltip_cols, [])
 
-    data_copy = data.loc[:, r_[cols, tt_cols]].copy()
-    data_copy.loc[:, cols] = data_copy.loc[:, cols].isna()
+    data_copy = data.loc[:, cols].copy().isna()
     if sort:
-        cols_sorted = data_copy.loc[:, cols]\
+        cols_sorted = data_copy\
             .sum()\
             .sort_values(ascending=False)\
             .index.tolist()
         data_copy.sort_values(by=cols_sorted, inplace=True)
         x_kws.update({'sort': cols_sorted})
 
     if droppable:
-        non_na_mask = ~data_copy.loc[:, cols].values
-        na_rows_mask = data_copy.loc[:, cols].any(axis=1).values[:, None]
+        non_na_mask = ~data_copy.values
+        na_rows_mask = data_copy.any(
+            axis=1).values[:, None]
         droppable_mask = non_na_mask & na_rows_mask
-        data_copy.loc[:, cols] = data_copy.loc[:, cols].astype(int)
-        data_copy.loc[:, cols] = data_copy.loc[:, cols]\
+        data_copy = data_copy.astype(int)\
             .mask(droppable_mask, other=2)
     else:
-        data_copy.loc[:, cols] = data_copy.loc[:, cols].astype(int)
+        data_copy = data_copy.astype(int)
 
-    data_copy.loc[:, cols] = data_copy.loc[:, cols].replace(
+    data_copy = data_copy.replace(
         dict(zip([0, 1, 2], names)))
 
     data_copy[ylabel] = arange(data.shape[0])
     data_copy = data_copy.melt(
-        id_vars=r_[[ylabel], tt_cols],
+        id_vars=[ylabel],
         value_vars=cols,
         var_name=xlabel,
         value_name=zlabel)
@@ -598,8 +593,7 @@ def plot_heatmap(
             x=X(**x_kws),
             y=Y(**y_kws),
             color=Color(**color_kws),
-            tooltip=tt_cols.tolist()
-        )
+    )
 
     return chart\
         .configure_axis(labelFontSize=font_size, titleFontSize=font_size)\