src/find_interaction_cluster/community_figures/*.py: add a paramter...

src/find_interaction_cluster/community_figures/*.py: add a paramter display_size to display the size of every community

src/find_interaction_cluster/community_figures/__main__.py

@@ -54,7 +54,8 @@ def load_and_check_table(table: str, feature: str, target_col: str):
 def create_community_figures(table: str, feature: str, target_col: str,
                              output: str, outfile: str, test_type: str,
                              target_kind: str = "", sd_community: str = "y",
-                             iteration: int = 10000) -> None:
+                             iteration: int = 10000,
+                             display_size: bool = False) -> None:
     """
     Create a dataframe with a control community, save it as a table and \
     as a barplot figure.
@@ -75,6 +76,8 @@ def create_community_figures(table: str, feature: str, target_col: str,
     is only used if test_type = 'permutation' (default 10000).
     :param sd_community: y to display the standard deviation for communities \
     false else.
+    :param display_size: True to display the size of the community above \
+    each one of them False to display nothing. (default False)
     """
     df = load_and_check_table(table, feature, target_col)
     if not outfile.endswith(".pdf"):
@@ -83,7 +86,7 @@ def create_community_figures(table: str, feature: str, target_col: str,
     sd_community = sd_community.lower() == 'y'
     create_community_fig(df, feature, target_col, moutfile, test_type,
                          target_kind=target_kind, iteration=iteration,
-                         sd_community=sd_community)
+                         sd_community=sd_community, display_size=display_size)
 
 
 if __name__ == "__main__":

src/find_interaction_cluster/community_figures/fig_functions.py

@@ -316,9 +316,41 @@ def expand_results_perm(df: pd.DataFrame, rdf: pd.DataFrame, target_col: str,
     return pd.concat([df_ctrl, df], axis=0, ignore_index=True)
 
 
+def display_size_fig(g: sns.FacetGrid, display_size: bool, target_col: str,
+                     df_bar: pd.DataFrame):
+    """
+
+    :param g: A seaborn FacetGrid
+    :param display_size: True to display the size of the community above \
+    each one of them False to display nothing. (default False)
+    :param target_col: The name of the column containing the data of interest
+    :param df_bar: A dataframe with the enrichment of a \
+    nucleotide frequency for every community (without control)
+    :return: The seaborn fascetgrid
+    """
+    xrange = g.ax.get_xlim()
+    if display_size:
+        df_size = df_bar[[f"mean_{target_col}", "community",
+                          "community_size"]].drop_duplicates()
+        ax2 = g.ax.twinx()
+        ax2.set_ylabel('community_size', color="green")
+        df_size.plot(x="community", y="community_size", kind="scatter", ax=ax2,
+                     legend=False, zorder=55,
+                     color=(0.2, 0.8, 0.2, 0.4))
+        ax2.tick_params(axis='y', labelcolor="green")
+        ax2.grid(False)
+        sizes = df_size["community_size"].to_list()
+        if max(sizes) - min(sizes) > 500:
+            ax2.set_yscale("log")
+            g.ax.set_xlim(xrange)
+    g.set(xticklabels=[])
+    return g
+
+
 def make_barplot(df_bar: pd.DataFrame, outfile: Path,
                  target_col: str, feature: str, target_kind: str = "",
-                 sd_community: Optional[str] = "sd") -> None:
+                 sd_community: Optional[str] = "sd",
+                 display_size: bool = False) -> None:
     """
     Create a barplot showing the frequency of `nt` for every community \
     of exons/gene in `df_bar`.
@@ -332,6 +364,8 @@ def make_barplot(df_bar: pd.DataFrame, outfile: Path,
     :param feature: The king of feature of interest
     :param sd_community: sd to display community error bar, None to display \
     nothing
+    :param display_size: True to display the size of the community above \
+    each one of them False to display nothing. (default False)
     """
     sns.set(context="poster")
     g = sns.catplot(x="community", y=target_col, data=df_bar, kind="point",
@@ -346,7 +380,7 @@ def make_barplot(df_bar: pd.DataFrame, outfile: Path,
     g.fig.suptitle(f"Mean frequency of {target_col}{target_kind}"
                    f"among community of {feature}s\n"
                    f"(stats obtained with a lm test)")
-    g.set(xticklabels=[])
+    g = display_size_fig(g, display_size, target_col, df_bar)
     g.ax.set_ylabel(f'Frequency of {target_col}')
     df_bara = df_bar.drop_duplicates(subset="community", keep="first")
     for i, p in enumerate(g2.ax.patches):
@@ -366,7 +400,8 @@ def make_barplot(df_bar: pd.DataFrame, outfile: Path,
 def make_barplot_perm(df_bar: pd.DataFrame, outfile: Path,
                       target_col: str, feature: str,
                       target_kind: str = "",
-                      sd_community: Optional[str] = "sd") -> None:
+                      sd_community: Optional[str] = "sd",
+                      display_size: bool = False) -> None:
     """
     Create a barplot showing the frequency of `nt` for every community \
     of exons/gene in `df_bar`.
@@ -380,6 +415,8 @@ def make_barplot_perm(df_bar: pd.DataFrame, outfile: Path,
     :param feature: The king of feature of interest
     :param sd_community: sd to display community error bar, None to display \
     nothing
+    :param display_size: True to display the size of the community above \
+    each one of them False to display nothing. (default False)
     """
     sns.set(context="poster")
     df_ctrl = df_bar.loc[df_bar[f"id_{feature}"] == 'ctrl', :]
@@ -404,7 +441,7 @@ def make_barplot_perm(df_bar: pd.DataFrame, outfile: Path,
     g.fig.suptitle(f"Mean frequency of {target_col}{target_kind}"
                    f"among community of {feature}s\n"
                    f"(stats obtained with a permutation test)")
-    g.set(xticklabels=[])
+    g = display_size_fig(g, display_size, target_col, df_bar)
     g.ax.set_ylabel(f'Frequency of {target_col}')
     df_bara = df_bar.drop_duplicates(subset="community", keep="first")
     for i, p in enumerate(g2.ax.patches):
@@ -423,7 +460,8 @@ def make_barplot_perm(df_bar: pd.DataFrame, outfile: Path,
 
 def barplot_creation(df_bar: pd.DataFrame, outfig: Path,
                      cpnt: str, test_type: str, feature: str,
-                     target_kind: str, sd_community: bool) -> None:
+                     target_kind: str, sd_community: bool,
+                     display_size: bool) -> None:
     """
     Reformat a dataframe with the enrichment of a nucleotide frequency \
     for every feature for every community and then create a \
@@ -441,13 +479,16 @@ def barplot_creation(df_bar: pd.DataFrame, outfig: Path,
     target_col.
     :param sd_community: True to display the errors bars for communities,
     False else.
+    :param display_size: True to display the size of the community above \
+    each one of them False to display nothing. (default False)
     """
     sd_community = "sd" if sd_community else None
     if test_type == "lm":
-        make_barplot(df_bar, outfig, cpnt, feature, target_kind)
+        make_barplot(df_bar, outfig, cpnt, feature, target_kind,
+                     display_size=display_size)
     else:
         make_barplot_perm(df_bar, outfig, cpnt, feature, target_kind,
-                          sd_community)
+                          sd_community, display_size)
 
 
 def get_feature_by_community(df: pd.DataFrame, feature: str) -> Dict:
@@ -486,7 +527,8 @@ def create_community_fig(df: pd.DataFrame, feature: str,
                          dic_com: Optional[Dict] = None,
                          target_kind: str = "",
                          iteration: int = 10000,
-                         sd_community: bool = True) -> None:
+                         sd_community: bool = True,
+                         display_size: bool = False) -> None:
     """
     Create a dataframe with a control community, save it as a table and \
     as a barplot figure.
@@ -507,6 +549,8 @@ def create_community_fig(df: pd.DataFrame, feature: str,
     :param iteration: The number of sub samples to create
     :param sd_community: True to display the errors bars for communities,
     False else.
+    :param display_size: True to display the size of the community above \
+    each one of them False to display nothing. (default False)
     """
     if dic_com is None:
         dic_com = {} if test_type == 'lm' \
@@ -522,4 +566,4 @@ def create_community_fig(df: pd.DataFrame, feature: str,
     bar_outfile = str(outfile_ctrl).replace(".pdf", "_bar.txt")
     df_bar.to_csv(bar_outfile, sep="\t", index=False)
     barplot_creation(df_bar, outfile_ctrl, target_col, test_type, feature,
-                     target_kind, sd_community)
+                     target_kind, sd_community, display_size)