Fix Plots

- add missing xlabel
- change "Mean Difference" to "Mean Distance"
- reduce plotting data
 - plot max 1000000 random positions to reduce runtime and prevent crashing

Author: JannesSP

magnipore/magnipore.py

@@ -244,10 +244,10 @@ def magnipore(mapping : dict, unaligned : dict, seq_dict : dict, aln_dict: dict,
     num_indels, sign_pos, nans = 0, 0, 0
 
     # TODO add some quality value
-    plotting_data = pd.DataFrame(columns=['Mean Difference', 'Avg Stdev', 'Strand', 'Mutational Context', 'Significant', 'TD Score', 'KL Divergence'])
+    plotting_data = pd.DataFrame(columns=['Mean Distance', 'Avg Stdev', 'Strand', 'Mutational Context', 'Significant', 'TD Score', 'KL Divergence'])
     plotting_data = plotting_data.astype(
         {
-            'Mean Difference': 'float32',
+            'Mean Distance': 'float32',
             'Avg Stdev': 'float32',
             'Strand': 'bool',
             'Mutational Context': 'bool',
@@ -309,7 +309,7 @@ def magnipore(mapping : dict, unaligned : dict, seq_dict : dict, aln_dict: dict,
                 td = td_score(mDiff, sAvg)
                 significant = td>=1
                 new_entry = pd.DataFrame({
-                            'Mean Difference' : [mDiff],
+                            'Mean Distance' : [mDiff],
                             'Avg Stdev' : [sAvg],
                             'Strand' : [strand],
                             'Mutational Context' : [mut_context],
@@ -385,10 +385,15 @@ def plotStatistics(plotting_data : pd.DataFrame, working_dir : str, first_sample
     plot_dir = os.path.join(working_dir, 'magnipore', f'{first_sample_label}_{sec_sample_label}', 'plots')
     if not os.path.exists(plot_dir):
         os.mkdir(plot_dir)
-    ### Mean Dist vs Std Avg plot
+    # reduce plotting_data, if it got too large too reduce runtime and prevent the kernel from killing the process
+    plotting_threshold = 1000000 # arbitrary threshold
+    if len(plotting_data.index) > plotting_threshold:
+        LOGGER.printLog(f'The number of positions exceeds the threshold of {plotting_data} ({len(plotting_data.index)}). To prevent the kernel from killing the process, Magnipore will only plot a subset of {plotting_data} positions. Plots will not include the full data.')
+        plotting_data = plotting_data.sample(plotting_threshold, replace=False)
+    # Mean Dist vs Std Avg plot
     LOGGER.printLog('Plotting Mean vs Stdev')
     plotMeanDiffStdAvg(plotting_data, plot_dir, first_sample_label, sec_sample_label)
-    ### plot scores
+    # plot scores
     LOGGER.printLog(f'Plotting TD score and KL divergence')
     plotScores(plotting_data, plot_dir, first_sample_label, sec_sample_label)
 
@@ -434,17 +439,17 @@ def plotMeanDiffStdAvg(dataframe : pd.DataFrame, working_dir : str, first_sample
     label1 = first_sample_label.replace("_", " ")
     label2 = sec_sample_label.replace("_", " ")
 
-    g = sns.JointGrid(x='Mean Difference', y='Avg Stdev', data=dataframe, hue='Mutational Context', marginal_ticks=True, palette=['blue', '#d95f02'], hue_order=[True, False], height = 10)
+    g = sns.JointGrid(x='Mean Distance', y='Avg Stdev', data=dataframe, hue='Mutational Context', marginal_ticks=True, palette=['blue', '#d95f02'], hue_order=[True, False], height = 10)
     g.plot_joint(func=sns.scatterplot, s = 8)
     g.ax_joint.cla()
     for _, row in dataframe.iterrows():
-        g.ax_joint.plot(row['Mean Difference'], row['Avg Stdev'], color = color(row['Mutational Context']), marker = marker(row['Mutational Context']), markersize=3, alpha = 0.6)
+        g.ax_joint.plot(row['Mean Distance'], row['Avg Stdev'], color = color(row['Mutational Context']), marker = marker(row['Mutational Context']), markersize=3, alpha = 0.6)
 
-    g.fig.suptitle(f'{len(dataframe.index)} compared bases mean difference against\naverage standard deviation\n{label1} and {label2}', y=0.98)
+    g.fig.suptitle(f'{len(dataframe.index)} compared bases mean distance against\naverage standard deviation\n{label1} and {label2}', y=0.98)
     g.ax_joint.grid(True, 'both', 'both', alpha = 0.4, linestyle = '--', linewidth = 0.5)
 
     lims = np.array([
-        [-.02, max(dataframe['Mean Difference']) + 0.1],
+        [-.02, max(dataframe['Mean Distance']) + 0.1],
         [-.02, max(dataframe['Avg Stdev']) + 0.1]
     ])
 
@@ -457,7 +462,8 @@ def plotMeanDiffStdAvg(dataframe : pd.DataFrame, working_dir : str, first_sample
     g.ax_joint.set_xlim(tuple(lims[0]))
     g.ax_joint.set_ylim(tuple(lims[1]))
 
-    g.ax_joint.set_ylabel('Average Standard Deviation')
+    g.ax_joint.set_ylabel('Average standard deviation')
+    g.ax_joint.set_xlabel('Mean distance')
     legend_mut = mlines.Line2D([], [], color='blue', marker='D', linestyle='None', markersize=10, label='mutation')
     legend_mod = mlines.Line2D([], [], color='#d95f02', marker='o', linestyle='None', markersize=10, label='matching reference')
     sign = mlines.Line2D([], [], color='#1b9e77', marker='s', linestyle='None', markersize=10, label='significant, TD>=1')