minor fix

examples/pose_centroids_molplotly.ipynb

@@ -614,6 +614,30 @@
  20.56523380975767
  ],
  "yaxis": "y"
+ },
+ {
+ "marker": {
+ "color": "red",
+ "size": 12,
+ "symbol": "x"
+ },
+ "mode": "markers",
+ "showlegend": false,
+ "type": "scatter",
+ "x": [
+ 20.710113587648642,
+ 4.312267729434556,
+ 18.935473315914678,
+ 17.68965247025577,
+ 20.61053154671133
+ ],
+ "y": [
+ 20.431212351814835,
+ 4.189062767055896,
+ 18.95856209889753,
+ 17.674840683395807,
+ 20.55944905349298
+ ]
  }
  ],
  "layout": {
@@ -1473,7 +1497,7 @@
  }
  ],
  "source": [
- "import plotly.express as px\n",
+ "import plotly.graph_objects as go\n",
  "\n",
  "rmsd_df['pose_index'] = range(1, len(renamed_poses)+1)\n",
  "rmsd_df['cluster'] = kmeans.labels_\n",
@@ -1490,14 +1514,25 @@
  " labels={'1':'RMSD to pose 1',\n",
  " '2':'RMSD to pose 2',}\n",
  " )\n",
+ "scatter_fig.add_trace(\n",
+ " go.Scatter(\n",
+ " x = kmeans.cluster_centers_[:,0],\n",
+ " y = kmeans.cluster_centers_[:,1],\n",
+ " mode='markers',\n",
+ " marker=dict(color=\"red\", \n",
+ " symbol='x',\n",
+ " size=10),\n",
+ " showlegend=False,))\n",
  "scatter_fig.update_traces(marker=dict(size=12))\n"
  ]
  },
  {
  "cell_type": "markdown",
  "metadata": {},
  "source": [
- "Adding `molplotly` and specifying `mol_col` allows the 3D coordinates of the pose to be shown on mouseover!"
+ "Adding `molplotly` and specifying `mol_col` allows the 3D coordinates of the pose to be shown on mouseover!\n",
+ "\n",
+ "Unfortunately there are issues with handling additional non-molecular traces in `molplotly` so we'll have to re-create the scatter plot without the centroids."
  ]
  },
  {
@@ -1520,7 +1555,7 @@
  " "
  ],
  "text/plain": [
- "<IPython.lib.display.IFrame at 0x13733e8c0>"
+ "<IPython.lib.display.IFrame at 0x139627d30>"
  ]
  },
  "metadata": {},
@@ -1532,6 +1567,20 @@
  "\n",
  "rmsd_df['mol'] = renamed_poses \n",
  "\n",
+ "scatter_fig = px.scatter(rmsd_df, \n",
+ " x=\"1\", \n",
+ " y=\"2\",\n",
+ " color='cluster',\n",
+ " hover_name='pose_index',\n",
+ " width=1000,\n",
+ " height=800,\n",
+ " title='Clustering of ligand poses',\n",
+ " labels={'1':'RMSD to pose 1',\n",
+ " '2':'RMSD to pose 2',}\n",
+ " )\n",
+ "\n",
+ "scatter_fig.update_traces(marker=dict(size=12))\n",
+ "\n",
  "app_clusters = molplotly.add_molecules(\n",
  " fig=scatter_fig,\n",
  " df=rmsd_df,\n",
@@ -1553,7 +1602,7 @@
  },
  {
  "cell_type": "code",
- "execution_count": 7,
+ "execution_count": 6,
  "metadata": {},
  "outputs": [
  {
@@ -1571,7 +1620,7 @@
  " "
  ],
  "text/plain": [
- "<IPython.lib.display.IFrame at 0x13733f820>"
+ "<IPython.lib.display.IFrame at 0x13969be50>"
  ]
  },
  "metadata": {},

molplotly/main.py

@@ -55,9 +55,10 @@ def find_grouping(
 
  if fig.data[0].hovertemplate is not None:
  col_names = re.findall(r"(.*?)=(?!%).*?<.*?>", fig.data[0].hovertemplate)
+ col_names = [re.sub(r"(.*)>", "", col_name) for col_name in col_names]
  if set(col_names) != set(cols):
  raise ValueError(
- "marker_col/color_col/facet_col is misspecified because the dataframe grouping names don't match the names in the plotly figure.",
+ f"marker_col/color_col/facet_col is misspecified because the specified dataframe grouping names {cols} don't match the names in the plotly figure {col_names}.",
  )
 
  df_grouped = df_data.groupby(col_names)