[Jungfrau][Correct] Avoid `NAN` disturbance for correction plots scale

Description

As clear from the plots while I have been looking through different plots from the latest corrections. I see JF4M has bad-scaled corrected plots. This is because of the NANs in the corrected data.

How Has This Been Tested?

Relevant Documents (optional)

jf4m_master.pdf

Types of changes

• Bug fix (non-breaking change which fixes an issue)

Checklist:

Reviewers

@schmidtp @kluyvert

notebooks/Jungfrau/Jungfrau_Gain_Correct_and_Verify_NBC.ipynb

@@ -775,6 +775,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
+    "step_timer.start()\n",
     "first_seq = 0 if sequences == [-1] else sequences[0]\n",
     "\n",
     "corrected_files = [\n",
@@ -817,7 +818,8 @@
     "gain = jf_raw.get_array(\"data.gain\")[:, :, cell_idx_preview, ...].values\n",
     "gain_train_cells = (\n",
     "    jf_raw.select_trains(by_id[[tid]]).get_array(\"data.gain\")[:, :, :, ...].values\n",
-    ")"
+    ")\n",
+    "step_timer.done_step(\"Prepared data for plotting\")"
    ]
   },
   {
@@ -865,8 +867,8 @@
     "print(f\"The per pixel mean of the first {corrected.shape[1]} trains of the first sequence file\")\n",
     "\n",
     "fig, ax = plt.subplots(figsize=(18, 10))\n",
-    "corrected_mean = np.mean(corrected, axis=1)\n",
-    "vmin, vmax = np.percentile(corrected_mean, [5, 95])\n",
+    "corrected_mean = np.nanmean(corrected, axis=1)\n",
+    "vmin, vmax = np.nanpercentile(corrected_mean, [5, 95])\n",
     "\n",
     "mean_plot_kwargs = dict(vmin=vmin, vmax=vmax)\n",
     "\n",
@@ -929,7 +931,7 @@
     "display(Markdown((f\"#### A single image from train {tid}\")))\n",
     "\n",
     "fig, ax = plt.subplots(figsize=(18, 10))\n",
-    "vmin, vmax = np.percentile(corrected_train, [5, 95])\n",
+    "vmin, vmax = np.nanpercentile(corrected_train, [5, 95])\n",
     "\n",
     "single_plot_kwargs = dict(\n",
     "    vmin=vmin,\n",