feat: select columns for automatic analysis

renumics/spotlight/analysis/__init__.py

@@ -19,7 +19,9 @@
     importlib.import_module(analyzers_namespace.__name__ + "." + module_info.name)
 
 
-def find_issues(data_source: DataSource, dtypes: ColumnTypeMapping) -> List[DataIssue]:
+def find_issues(
+    data_source: DataSource, columns: List[str], dtypes: ColumnTypeMapping
+) -> List[DataIssue]:
     """
     Find dataset issues in the data source
     """
@@ -28,7 +30,7 @@ def find_issues(data_source: DataSource, dtypes: ColumnTypeMapping) -> List[Data
 
     issues: List[DataIssue] = []
     for analyze in registered_analyzers:
-        issues.extend(analyze(data_source, dtypes))
+        issues.extend(analyze(data_source, columns, dtypes))
 
     logger.info("Analysis done.")
 

renumics/spotlight/analysis/analyzers/cleanlab.py

@@ -3,7 +3,7 @@
 """
 
 import inspect
-from typing import Iterable
+from typing import Iterable, List
 
 import numpy as np
 import cleanlab.outlier
@@ -19,13 +19,13 @@
 
 @data_analyzer
 def analyze_with_cleanlab(
-    data_source: DataSource, dtypes: ColumnTypeMapping
+    data_source: DataSource, columns: List[str], dtypes: ColumnTypeMapping
 ) -> Iterable[DataIssue]:
     """
     Find (embedding) outliers with cleanlab
     """
 
-    embedding_columns = (col for col, dtype in dtypes.items() if dtype == Embedding)
+    embedding_columns = (col for col in columns if dtypes.get(col) == Embedding)
     for column_name in embedding_columns:
         col_values = data_source.get_column(column_name, dtypes[column_name]).values
         embeddings = np.array(col_values, dtype=object)

renumics/spotlight/analysis/analyzers/cleanvision.py

@@ -103,13 +103,13 @@ def _get_cell_data_safe(
 
 @data_analyzer
 def analyze_with_cleanvision(
-    data_source: DataSource, dtypes: ColumnTypeMapping
+    data_source: DataSource, columns: List[str], dtypes: ColumnTypeMapping
 ) -> Iterable[DataIssue]:
     """
     find image issues using cleanvision
     """
 
-    image_columns = [col for col, dtype in dtypes.items() if dtype == Image]
+    image_columns = [col for col in columns if dtypes.get(col) == Image]
 
     for column_name in image_columns:
         # load image data from data source

renumics/spotlight/analysis/typing.py

@@ -22,4 +22,4 @@ class DataIssue(BaseModel):
     description: str = ""
 
 
-DataAnalyzer = Callable[[DataSource, ColumnTypeMapping], Iterable[DataIssue]]
+DataAnalyzer = Callable[[DataSource, List[str], ColumnTypeMapping], Iterable[DataIssue]]

renumics/spotlight/app.py

@@ -99,7 +99,7 @@ class SpotlightApp(FastAPI):
     # data issues
     issues: Optional[List[DataIssue]] = []
     _custom_issues: List[DataIssue] = []
-    analyze_issues: bool = True
+    analyze_columns: Union[List[str], bool] = False
 
     def __init__(self) -> None:
         super().__init__()
@@ -112,7 +112,7 @@ def __init__(self) -> None:
         self.vite_url = None
         self.username = ""
         self.filebrowsing_allowed = False
-        self.analyze_issues = False
+        self.analyze_columns = False
         self.issues = None
         self._custom_issues = []
 
@@ -303,7 +303,7 @@ def update(self, config: AppConfig) -> None:
         if config.dtypes is not None:
             self._user_dtypes = config.dtypes
         if config.analyze is not None:
-            self.analyze_issues = config.analyze
+            self.analyze_columns = config.analyze
         if config.custom_issues is not None:
             self.custom_issues = config.custom_issues
         if config.dataset is not None:
@@ -407,7 +407,8 @@ def _update_issues(self) -> None:
         Update issues and notify client about.
         """
 
-        if not self.analyze_issues:
+        # early out for [] and False
+        if not self.analyze_columns:
             self.issues = []
             self._broadcast(IssuesUpdatedMessage())
             return
@@ -417,8 +418,14 @@ def _update_issues(self) -> None:
         self._broadcast(IssuesUpdatedMessage())
         if table is None:
             return
+
+        if self.analyze_columns is True:
+            columns = table.column_names
+        else:
+            columns = self.analyze_columns
+
         task = self.task_manager.create_task(
-            find_issues, (table, self._dtypes), name="update_issues"
+            find_issues, (table, columns, self._dtypes), name="update_issues"
         )
 
         def _on_issues_ready(future: Future) -> None:

renumics/spotlight/app_config.py

@@ -27,7 +27,7 @@ class AppConfig:
     project_root: Optional[Path] = None
 
     # data analysis
-    analyze: Optional[bool] = None
+    analyze: Optional[Union[bool, List[str]]] = None
     custom_issues: Optional[List[DataIssue]] = None
 
     # frontend

renumics/spotlight/cli.py

@@ -6,7 +6,7 @@
 import platform
 import signal
 import sys
-from typing import Optional, Tuple, Union
+from typing import Optional, Tuple, Union, List
 from pathlib import Path
 
 import click
@@ -95,10 +95,15 @@ def cli_dtype_callback(
     help="Whether to allow users to browse and open datasets.",
 )
 @click.option(
-    "--analyze",
+    "--analyze-all",
     is_flag=True,
     default=False,
-    help="Automatically analyze common dataset errors.",
+    help="Automatically analyze issues for all columns.",
+)
+@click.option(
+    "--analyze",
+    default=[],
+    help="Automatically analyze issues for all columns.",
 )
 @click.option("-v", "--verbose", is_flag=True)
 @click.version_option(spotlight.__version__)
@@ -110,7 +115,8 @@ def main(
     dtype: Optional[ColumnTypeMapping],
     no_browser: bool,
     filebrowsing: bool,
-    analyze: bool,
+    analyze: List[str],
+    analyze_all: bool,
     verbose: bool,
 ) -> None:
     """
@@ -134,5 +140,5 @@ def main(
         no_browser=no_browser,
         allow_filebrowsing=filebrowsing,
         wait="forever",
-        analyze=analyze,
+        analyze=True if analyze_all else analyze,
     )

renumics/spotlight/viewer.py

@@ -107,7 +107,7 @@ def show(
         allow_filebrowsing: Union[bool, Literal["auto"]] = "auto",
         wait: Union[bool, Literal["auto", "forever"]] = "auto",
         dtype: Optional[ColumnTypeMapping] = None,
-        analyze: Optional[bool] = None,
+        analyze: Optional[Union[bool, List[str]]] = None,
         issues: Optional[Collection[DataIssue]] = None,
     ) -> None:
         """
@@ -320,7 +320,7 @@ def show(
     allow_filebrowsing: Union[bool, Literal["auto"]] = "auto",
     wait: Union[bool, Literal["auto", "forever"]] = "auto",
     dtype: Optional[ColumnTypeMapping] = None,
-    analyze: Optional[bool] = None,
+    analyze: Optional[Union[bool, List[str]]] = None,
     issues: Optional[Collection[DataIssue]] = None,
 ) -> Viewer:
     """