fix(eda): fixed uncaught dtype exceptions

sfu-db · Apr 26, 2021 · d301eb7 · d301eb7
1 parent 9cb11f4
commit d301eb7
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 7 deletions.
diff --git a/dataprep/eda/create_report/formatter.py b/dataprep/eda/create_report/formatter.py
@@ -30,6 +30,7 @@
     Continuous,
     DateTime,
     Nominal,
+    GeoGraphy,
     detect_dtype,
     is_dtype,
 )
@@ -127,6 +128,8 @@ def format_basic(df: dd.DataFrame, cfg: Config) -> Dict[str, Any]:
                 ins = _format_cont_ins(col, dat, data["ov"]["nrows"], cfg)[1]
             elif is_dtype(dtp, Nominal()):
                 ins = _format_nom_ins(col, dat, data["ov"]["nrows"], cfg)[1]
+            elif is_dtype(dtp, GeoGraphy()):
+                ins = _format_nom_ins(col, dat, data["ov"]["nrows"], cfg)[1]
             else:
                 continue
             all_ins += ins
@@ -150,6 +153,11 @@ def format_basic(df: dd.DataFrame, cfg: Config) -> Dict[str, Any]:
                 stats = format_cat_stats(
                     data[col]["stats"], data[col]["len_stats"], data[col]["letter_stats"]
                 )
+            elif is_dtype(detect_dtype(df[col]), GeoGraphy()):
+                itmdt = Intermediate(col=col, data=data[col], visual_type="categorical_column")
+                stats = format_cat_stats(
+                    data[col]["stats"], data[col]["len_stats"], data[col]["letter_stats"]
+                )
             elif is_dtype(detect_dtype(df[col]), DateTime()):
                 itmdt = Intermediate(
                     col=col,
@@ -268,6 +276,8 @@ def basic_computations(
                 data[col] = nom_comps(df.frame[col], df.frame[col].head(), cfg)
             elif is_dtype(detect_dtype(df.frame[col]), Nominal()):
                 data[col] = nom_comps(df.frame[col], first_rows[col], cfg)
+            elif is_dtype(detect_dtype(df.frame[col]), GeoGraphy()):
+                data[col] = nom_comps(df.frame[col], first_rows[col], cfg)
             elif is_dtype(detect_dtype(df.frame[col]), Continuous()):
                 data[col] = cont_comps(df.frame[col], cfg)
             elif is_dtype(detect_dtype(df.frame[col]), DateTime()):
@@ -289,6 +299,10 @@ def basic_computations(
                 data["insights"].append(
                     (col, Nominal(), _nom_calcs(df.frame[col].dropna(), head[col], cfg))
                 )
+            elif is_dtype(col_dtype, GeoGraphy()):
+                data["insights"].append(
+                    (col, Nominal(), _nom_calcs(df.frame[col].dropna(), head[col], cfg))
+                )
             elif is_dtype(col_dtype, DateTime()):
                 data["insights"].append(
                     (col, DateTime(), dask.delayed(_calc_line_dt)(df.frame[[col]], cfg.line.unit))

diff --git a/dataprep/eda/create_report/templates/scripts.html b/dataprep/eda/create_report/templates/scripts.html
@@ -29,19 +29,17 @@
         for (let i of element) {
             let thValue = i.firstElementChild.innerText;
             let tbValue = i.lastElementChild.innerText;
-            if (thValue.includes('Distinct') && tbValue > 50) {
-                i.style.color = color;
-            } else if (thValue.includes('Unique') && tbValue.replace('%',
-                    '') == 100) {
+            if (thValue.includes('Unique') && tbValue.replace('%',
+                '') == 100) {
                 i.style.color = color;
             } else if (thValue.includes('Missing') && tbValue.replace('%',
-                    '') != 0) {
+                '') != 0) {
                 i.style.color = color;
             } else if (thValue.includes('Zeros') && tbValue.replace('%',
-                    '') != 0) {
+                '') != 0) {
                 i.style.color = color;
             } else if (thValue.includes('Infinite') && tbValue.replace('%',
-                    '') != 0) {
+                '') != 0) {
                 i.style.color = color;
             } else if (thValue.includes('Skewness' && tbValue > 20)) {
                 i.style.color = color;