updates

modules/10-spatial-models/lecture.ipynb

@@ -83,7 +83,8 @@
     "import mgwr\n",
     "import pandas as pd\n",
     "import spreg\n",
-    "from libpysal import weights"
+    "from libpysal import weights\n",
+    "from matplotlib import colormaps as cmaps"
    ]
   },
   {
@@ -112,8 +113,8 @@
    "outputs": [],
    "source": [
     "# load CA tract-level census variables\n",
-    "df_census = pd.read_csv(\"../../data/census_tracts_data_ca.csv\", dtype={\"GEOID10\": str}).set_index(\n",
-    "    \"GEOID10\"\n",
+    "df_census = pd.read_csv(\n",
+    "    \"../../data/census_tracts_data_ca.csv\", dtype={\"GEOID10\": str}, index_col=\"GEOID10\"\n",
     ")\n",
     "df_census.shape"
    ]
@@ -452,7 +453,7 @@
     "    from matplotlib.colors import LinearSegmentedColormap as lsc\n",
     "\n",
     "    name = f\"{cmap_name}_new\"\n",
-    "    cmap = plt.cm.get_cmap(cmap_name)\n",
+    "    cmap = cmaps.get_cmap(cmap_name)\n",
     "    vmin = values.min()\n",
     "    vmax = values.max()\n",
     "\n",
@@ -579,7 +580,7 @@
    "outputs": [],
    "source": [
     "# compute spatial weights for only those tracts that appear in design matrix\n",
-    "W = weights.Queen.from_dataframe(tracts.loc[X.index])\n",
+    "W = weights.Queen.from_dataframe(tracts.loc[X.index], use_index=True)\n",
     "W.transform = \"r\""
    ]
   },
@@ -735,7 +736,7 @@
     "\n",
     "> Due to spatial spillover, each coefficient alone does not represent the marginal effect on the response of a unit increase in the predictor. Instead, it represents the direct effect: what happens locally if you make a unit change in the predictor only in one tract. But also present are indirect effects: local spillovers in each tract from a unit predictor change in other tracts.\n",
     "\n",
-    "Refer to the article for details on how to calculate and interpret total effects.\n",
+    "Refer to the article for details on how to calculate and interpret *total effects*.\n",
     "\n",
     "## 8. Spatial error model\n",
     "\n",