Add embed argument to spotlight.show and CLI

renumics/spotlight/app.py

@@ -123,7 +123,10 @@ class SpotlightApp(FastAPI):
     # data issues
     issues: Optional[List[DataIssue]] = []
     _custom_issues: List[DataIssue] = []
-    analyze_columns: Union[List[str], bool] = False
+    analyze_columns: Union[List[str], bool]
+
+    # embedding
+    embed_columns: Union[List[str], bool]
 
     def __init__(self) -> None:
         super().__init__()
@@ -138,6 +141,7 @@ def __init__(self) -> None:
         self.analyze_columns = False
         self.issues = None
         self._custom_issues = []
+        self.embed_columns = False
 
         self._dataset = None
         self._user_dtypes = {}
@@ -328,6 +332,8 @@ def update(self, config: AppConfig) -> None:
                 self.analyze_columns = config.analyze
             if config.custom_issues is not None:
                 self.custom_issues = config.custom_issues
+            if config.embed is not None:
+                self.embed_columns = config.embed
             if config.dataset is not None:
                 self._dataset = config.dataset
                 self._data_source = create_datasource(self._dataset)
@@ -464,12 +470,24 @@ def _update_embeddings(self) -> None:
         """
         Update embeddings, update them in the data store and notify client about.
         """
+        if not self.embed_columns:
+            return
+
         if self._data_store is None:
             return
 
         logger.info("Embedding started.")
 
-        embedders = create_embedders(self._data_store, self._data_store.column_names)
+        if self.embed_columns is True:
+            embed_columns = self._data_store.column_names
+        else:
+            embed_columns = [
+                column
+                for column in self.embed_columns
+                if column in self._data_store.column_names
+            ]
+
+        embedders = create_embedders(self._data_store, embed_columns)
 
         self._data_store.embeddings = {column: None for column in embedders}
 

renumics/spotlight/app_config.py

@@ -24,9 +24,12 @@ class AppConfig:
     project_root: Optional[Path] = None
 
     # data analysis
-    analyze: Optional[Union[bool, List[str]]] = None
+    analyze: Optional[Union[List[str], bool]] = None
     custom_issues: Optional[List[DataIssue]] = None
 
+    # embedding
+    embed: Optional[Union[List[str], bool]] = None
+
     # frontend
     layout: Optional[Layout] = None
     filebrowsing_allowed: Optional[bool] = None

renumics/spotlight/cli.py

@@ -6,7 +6,7 @@
 import platform
 import signal
 import sys
-from typing import Dict, Optional, Tuple, Union, List
+from typing import Dict, Optional, Tuple, Union
 
 import click
 
@@ -94,9 +94,21 @@ def cli_dtype_callback(
 )
 @click.option(
     "--analyze",
-    default=[],
+    default=(),
     multiple=True,
-    help="Automatically analyze issues for all columns.",
+    help="Columns to analyze (if no --analyze-all).",
+)
+@click.option(
+    "--embed-all",
+    is_flag=True,
+    default=False,
+    help="Automatically embed all columns.",
+)
+@click.option(
+    "--embed",
+    default=(),
+    multiple=True,
+    help="Columns to embed (if no --analyze-all).",
 )
 @click.option("-v", "--verbose", is_flag=True)
 @click.version_option(spotlight.__version__)
@@ -109,8 +121,10 @@ def main(
     dtype: Optional[Dict[str, str]],
     no_browser: bool,
     filebrowsing: bool,
-    analyze: List[str],
+    analyze: Tuple[str],
     analyze_all: bool,
+    embed: Tuple[str],
+    embed_all: bool,
     verbose: bool,
 ) -> None:
     """
@@ -135,5 +149,6 @@ def main(
         no_browser=no_browser,
         allow_filebrowsing=filebrowsing,
         wait="forever",
-        analyze=True if analyze_all else analyze,
+        analyze=True if analyze_all else list(analyze),
+        embed=True if embed_all else list(embed),
     )

renumics/spotlight/viewer.py

@@ -150,8 +150,9 @@ def show(
         allow_filebrowsing: Union[bool, Literal["auto"]] = "auto",
         wait: Union[bool, Literal["auto", "forever"]] = "auto",
         dtype: Optional[Dict[str, Any]] = None,
-        analyze: Optional[Union[bool, List[str]]] = None,
+        analyze: Optional[Union[List[str], bool]] = None,
         issues: Optional[Collection[DataIssue]] = None,
+        embed: Optional[Union[List[str], bool]] = None,
     ) -> None:
         """
         Show a dataset or folder in this spotlight viewer.
@@ -173,6 +174,8 @@ def show(
                 column types allowed by Spotlight (for dataframes only).
             analyze: Automatically analyze common dataset issues (disabled by default).
             issues: Custom dataset issues displayed in the viewer.
+            embed: Automatically embed all or given columns with default
+                embedders (disabled by default).
         """
 
         if is_pathtype(dataset):
@@ -206,6 +209,7 @@ def show(
             project_root=project_root,
             analyze=analyze,
             custom_issues=list(issues) if issues else None,
+            embed=embed,
             layout=parsed_layout,
             filebrowsing_allowed=filebrowsing_allowed,
         )
@@ -373,6 +377,7 @@ def show(
     dtype: Optional[Dict[str, Any]] = None,
     analyze: Optional[Union[bool, List[str]]] = None,
     issues: Optional[Collection[DataIssue]] = None,
+    embed: Optional[Union[List[str], bool]] = None,
 ) -> Viewer:
     """
     Start a new Spotlight viewer.
@@ -397,6 +402,8 @@ def show(
             column types allowed by Spotlight (for dataframes only).
         analyze: Automatically analyze common dataset issues (disabled by default).
         issues: Custom dataset issues displayed in the viewer.
+        embed: Automatically embed all or given columns with default
+            embedders (disabled by default).
     """
 
     viewer = None
@@ -419,6 +426,7 @@ def show(
         dtype=dtype,
         analyze=analyze,
         issues=issues,
+        embed=embed,
     )
     return viewer