add mean_abs and sum_abs to lineplot_and_heatmap (#183)

Allows display of site statistics reflecting the mean and sum of the absolute values of the mutation values.

Also, update GitHub actions and lint/format with newer versions of `black` and `ruff`.

.github/workflows/test.yaml

@@ -15,10 +15,10 @@ jobs:
     timeout-minutes: 60
     steps:
       - name: checkout
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
 
       - name: install python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         with:
           python-version: "3.11"
 

CHANGELOG.rst

@@ -6,6 +6,10 @@ All notable changes to this project will be documented in this file.
 
 The format is based on `Keep a Changelog <https://keepachangelog.com>`_.
 
+6.11
+----
+- In ``lineplot_and_heatmap``, add ``mean_abs`` and ``sum_abs`` as possible site statistics, reflecting the mean and sum of the absolute values of the mutation values.
+
 6.10
 ----
 - In ``lineplot_and_heatmap``, mutations that are hidden are by one slider filtered even if they fail other sliders. Addresses `this issue <https://github.com/dms-vep/dms-vep-pipeline-3/issues/96>`_

polyclonal/__init__.py

@@ -31,7 +31,7 @@
 
 __author__ = "`the Bloom lab <https://research.fhcrc.org/bloom/en.html>`_"
 __email__ = "[email protected]"
-__version__ = "6.10"
+__version__ = "6.11"
 __url__ = "https://github.com/jbloomlab/polyclonal"
 
 from polyclonal.alphabets import AAS

polyclonal/plot.py

@@ -520,7 +520,7 @@ def lineplot_and_heatmap(
         they fail other filters in `addtl_slider_stats`.
     init_floor_at_zero : bool
         Initial value for option to put floor of zero on value is `stat_col`.
-    init_site_statistic : {'sum', 'mean', 'max', 'min'}
+    init_site_statistic : {'sum', 'mean', 'max', 'min', 'mean_abs', 'sum_abs'}
         Initial value for site statistic in lineplot, calculated from `stat_col`.
     cell_size : float
         Size of cells in heatmap
@@ -594,6 +594,7 @@ def lineplot_and_heatmap(
     -------
     altair.Chart
         Interactive plot.
+
     """
     if addtl_tooltip_stats is None:
         addtl_tooltip_stats = []
@@ -978,7 +979,7 @@ def replace_std(col):
     )
 
     # make the site chart
-    site_statistics = ["sum", "mean", "max", "min"]
+    site_statistics = ["sum", "mean", "max", "min", "sum_abs", "mean_abs"]
     if init_site_statistic not in site_statistics:
         raise ValueError(f"invalid {init_site_statistic=}")
     if set(site_statistics).intersection(req_cols):
@@ -998,8 +999,15 @@ def replace_std(col):
         base_chart.transform_filter(
             (alt.datum.wildtype != alt.datum.mutant) & ~alt.datum["_stat_hide"]
         )
+        .transform_calculate(_stat_abs=alt.expr.abs(alt.datum["_stat"]))
         .transform_aggregate(
-            **{f"_stat_{stat}": f"{stat}(_stat)" for stat in site_statistics},
+            **{
+                f"_stat_{stat}": (
+                    stat.split("_")[0]
+                    + f"({'_stat_abs' if stat.endswith('_abs') else '_stat'})"
+                )
+                for stat in site_statistics
+            },
             groupby=[*site_prop_cols, category_col],
         )
         .transform_fold(

polyclonal/polyclonal.py

@@ -2867,6 +2867,7 @@ def filter_variants_by_seen_muts(
         variants_df : pandas.DataFrame
             Copy of input dataframe, with rows of variants
             that have unseen mutations removed.
+
         """
         variants_df = variants_df.copy()
 
@@ -3154,6 +3155,7 @@ def mut_escape_corr(self, ref_poly):
         -------
         corr_df : pandas.DataFrame
             Pairwise epitope correlations for escape.
+
         """
         if self.mut_escape_df is None or ref_poly.mut_escape_df is None:
             raise ValueError("Both objects must have `mut_escape_df` initialized.")

polyclonal/polyclonal_collection.py

@@ -851,6 +851,7 @@ def mut_escape_corr_heatmap(
             Plot squared correlation (eg, :math:`R^2` rather :math:`R`).
         **kwargs
             Keyword args for :func:`polyclonal.plot.corr_heatmap`
+
         """
         corr_label = {"pearson": "r", "kendall": "tau", "spearman": "rho"}[method]
         corr2_label = f"{corr_label}2"