Merge pull request #288 from DashAISoftware/feat/ux-dataset

Feat/ux dataset

Author: cristian-tamblay

DashAI/back/api/api_v1/endpoints/datasets.py

@@ -19,7 +19,6 @@
 from DashAI.back.dataloaders.classes.dashai_dataset import (
     get_columns_spec,
     get_dataset_info,
-    update_columns_spec,
 )
 from DashAI.back.dependencies.database.models import Dataset, Experiment
 
@@ -531,20 +530,16 @@ async def update_dataset(
     session_factory: sessionmaker = Depends(lambda: di["session_factory"]),
     config: Dict[str, Any] = Depends(lambda: di["config"]),
 ):
-    """Updates the name and/or task name of a dataset with the provided ID.
+    """Updates the name of a dataset with the provided ID.
 
     Parameters
     ----------
     dataset_id : int
         ID of the dataset to update.
     params : DatasetUpdateParams
         A dictionary containing the new values for the dataset.
-        name : str, optional
+        name : str
             New name for the dataset.
-        task_name : str, optional
-            New task name for the dataset.
-        columns : Dict[str, ColumnSpecItemParams], optional
-            New column specification for the dataset.
     session_factory : Callable[..., ContextManager[Session]]
         A factory that creates a context manager that handles a SQLAlchemy session.
         The generated session can be used to access and query the database.
@@ -557,12 +552,8 @@ async def update_dataset(
     with session_factory() as db:
         try:
             dataset = db.get(Dataset, dataset_id)
-            if params.columns:
-                update_columns_spec(f"{dataset.file_path}/dataset", params.columns)
-            elif params.name:
+            if params.name and params.name != dataset.name:
                 setattr(dataset, "name", params.name)
-                new_folder_path = config["DATASETS_PATH"] / params.name
-                os.rename(dataset.file_path, new_folder_path)
                 db.commit()
                 db.refresh(dataset)
                 return dataset

DashAI/back/api/api_v1/endpoints/notebook.py

@@ -69,6 +69,10 @@ def create_notebook(
             shutil.copytree(dataset_folder, new_folder_path, dirs_exist_ok=True)
 
             notebook_data = params.model_dump()
+            notebook_data = {
+                **notebook_data,
+                "name": notebook_data.get("name") or "Untitled Notebook",
+            }
             notebook_data["file_path"] = new_folder_path
             notebook_model = Notebook(**notebook_data)
             db.add(notebook_model)
@@ -338,3 +342,56 @@ async def create_dataset_from_notebook(
             ) from e
 
     return dataset
+
+
+@router.patch("/{notebook_id}")
+@inject
+async def update_notebook(
+    notebook_id: int,
+    params: schemas.NotebookUpdateParams,
+    session_factory: sessionmaker = Depends(lambda: di["session_factory"]),
+):
+    """Updates the name of a notebook with the provided ID.
+
+    Parameters
+    ----------
+    notebook_id : int
+        ID of the notebook to update.
+    params : NotebookUpdateParams
+        A dictionary containing the new values for the notebook.
+        name : str
+            New name for the notebook.
+    session_factory : Callable[..., ContextManager[Session]]
+        A factory that creates a context manager that handles a SQLAlchemy session.
+        The generated session can be used to access and query the database.
+
+    Returns
+    -------
+    Dict
+        A dictionary containing the updated dataset record.
+    """
+    with session_factory() as db:
+        try:
+            notebook = db.get(Notebook, notebook_id)
+            if not notebook:
+                raise HTTPException(
+                    status_code=status.HTTP_404_NOT_FOUND,
+                    detail="Notebook not found",
+                )
+            if not params.name or params.name == notebook.name:
+                raise HTTPException(
+                    status_code=status.HTTP_304_NOT_MODIFIED,
+                    detail="No fields to update",
+                )
+
+            setattr(notebook, "name", params.name)
+            db.commit()
+            db.refresh(notebook)
+        except Exception as e:
+            log.error(f"Error updating notebook {notebook_id}: {e}")
+            raise HTTPException(
+                status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
+                detail="Failed to update notebook",
+            ) from e
+
+    return notebook

DashAI/back/api/api_v1/schemas/converter_params.py

@@ -6,15 +6,15 @@
 class ConverterParams(BaseModel):
     order: int = 0
     params: Dict[str, Union[str, int, float, bool, None]] = None
-    scope: Dict[str, List[int]] = None
-    target_index: int = None
+    scope: Dict[str, Union[List[int], List[Dict[str, Any]]]] = None
+    target: Union[Dict[str, Any], None] = None
 
     def serialize(self) -> Dict[str, Any]:
         return {
             "order": self.order,
             "params": self.params,
             "scope": self.scope,
-            "target_index": self.target_index,
+            "target": self.target,
         }
 
 

DashAI/back/api/api_v1/schemas/datasets_params.py

@@ -22,7 +22,6 @@ class ColumnsSpecParams(BaseModel):
 
 class DatasetUpdateParams(BaseModel):
     name: str = None
-    columns: Dict[str, ColumnSpecItemParams] = None
 
 
 class DatasetUploadFromNotebookParams(BaseModel):

DashAI/back/api/api_v1/schemas/notebook_params.py

@@ -19,3 +19,7 @@ class Notebook(NotebookBase):
     created: datetime
     last_modified: datetime
     file_path: str
+
+
+class NotebookUpdateParams(BaseModel):
+    name: str = None

DashAI/back/job/converter_job.py

@@ -211,7 +211,14 @@ def instantiate_chain(
             # dataset to edit
             dataset_path = f"{converter_list.notebook.file_path}/dataset"
             loaded_dataset = load_dataset(dataset_path)
-            target_column_index = converter_list.parameters.pop("target_index")
+            print("Pre target column")
+            params = converter_list.parameters or {}
+            target_column_index = (
+                params["target"].get("idx")
+                if params.get("target") is not None
+                else None
+            )
+            print(target_column_index)
 
             if not loaded_dataset:
                 raise JobError(f"Dataset with path {dataset_path} not found")
@@ -348,7 +355,9 @@ def instantiate_chain(
                 converter_scope = converter_info["scope"]
 
                 # Process columns scope
-                columns_scope = [column - 1 for column in converter_scope["columns"]]
+                columns_scope = [
+                    column["idx"] - 1 for column in converter_scope["columns"]
+                ]
                 scope_column_indexes = sorted(set(columns_scope))
 
                 # If no columns specified, use all columns

DashAI/back/job/dataset_job.py

@@ -3,6 +3,7 @@
 import logging
 import os
 import shutil
+import uuid
 from typing import Any, Dict
 
 from kink import inject
@@ -57,15 +58,16 @@ async def run(
 
             parsed_params = parse_params(DatasetParams, json.dumps(params))
             dataloader = component_registry[parsed_params.dataloader]["class"]()
-            folder_path = config["DATASETS_PATH"] / parsed_params.name
+            random_name = str(uuid.uuid4())
+            folder_path = config["DATASETS_PATH"] / random_name
 
             try:
                 log.debug("Trying to create a new dataset path: %s", folder_path)
                 folder_path.mkdir(parents=True)
             except FileExistsError as e:
                 log.exception(e)
                 raise JobError(
-                    f"A dataset with the name {parsed_params.name} already exists."
+                    f"A dataset with the name {random_name} already exists."
                 ) from e
 
             try:
@@ -90,7 +92,11 @@ async def run(
                 try:
                     folder_path = os.path.realpath(folder_path)
                     new_dataset = Dataset(
-                        name=parsed_params.name,
+                        name=(
+                            parsed_params.name
+                            if parsed_params.name
+                            else "Untitled Dataset"
+                        ),
                         file_path=folder_path,
                     )
                     db.add(new_dataset)

DashAI/front/src/api/notebook.ts

@@ -53,3 +53,14 @@ export const createDatasetFromNotebook = async (
 
   return response.data;
 };
+
+export const updateNotebook = async (
+  id: number,
+  formData: object,
+): Promise<INotebook> => {
+  console.log("updating notebook with id:", id, "and formData:", formData);
+  const response = await api.patch(`${notebookEndpoint}/${id}`, {
+    ...formData,
+  });
+  return response.data;
+};

DashAI/front/src/components/notebooks/LeftBar.jsx

@@ -17,8 +17,10 @@ export default function DatasetsNotebooksBar({
   selectedNotebookId,
   onDatasetClick,
   onDatasetDelete,
+  onDatasetEdit,
   onNotebookClick,
   onNotebookDelete,
+  onNotebookEdit,
   handleNewSessionButton,
 }) {
   const [searchQuery, setSearchQuery] = useState("");
@@ -50,6 +52,25 @@ export default function DatasetsNotebooksBar({
     }
   };
 
+  const getDatasetDescription = (dataset) => {
+    return (
+      dataset.description ||
+      `${dataset.total_rows || 0} rows, ${dataset.total_columns || 0} cols`
+    );
+  };
+
+  const getNotebookDescription = (notebook) => {
+    if (notebook.dataset_id && datasets.length > 0) {
+      const associatedDataset = datasets.find(
+        (dataset) => dataset.id === notebook.dataset_id,
+      );
+      return associatedDataset?.name
+        ? `from ${associatedDataset.name} dataset`
+        : "No dataset";
+    }
+    return notebook.description || "";
+  };
+
   return (
     <SideBar>
       {/* Header */}
@@ -88,9 +109,11 @@ export default function DatasetsNotebooksBar({
           selectedItemId={selectedDatasetId}
           onItemClick={onDatasetClick}
           onItemDelete={onDatasetDelete}
+          onItemEdit={onDatasetEdit}
           defaultOpen={true}
           title="Available Datasets"
           Icon={StorageIcon}
+          getItemDescription={getDatasetDescription}
         />
 
         <Divider sx={{ width: "90%", bgcolor: "#252836", mx: "auto" }} />
@@ -100,10 +123,13 @@ export default function DatasetsNotebooksBar({
           selectedItemId={selectedNotebookId}
           onItemClick={onNotebookClick}
           onItemDelete={onNotebookDelete}
+          onItemEdit={onNotebookEdit}
           onItemInfo={handleNotebookInfo}
           defaultOpen={true}
           title="Notebooks"
           Icon={DescriptionIcon}
+          datasets={datasets}
+          getItemDescription={getNotebookDescription}
         />
       </Box>
 

DashAI/front/src/components/notebooks/NoteBox.jsx

@@ -0,0 +1,21 @@
+import { Box, Typography } from "@mui/material";
+
+export default function NoteBox({ message }) {
+  return (
+    <Box
+      sx={{
+        mt: 2,
+        p: 2,
+        bgcolor: "#212121",
+        borderRadius: 1,
+        border: "1px solid rgba(255, 255, 255, 0.1)",
+        mb: 2,
+      }}
+    >
+      <Typography variant="subtitle2" sx={{ color: "#00BEBB", mb: 1 }}>
+        Note:
+      </Typography>
+      <Typography variant="body2">{message}</Typography>
+    </Box>
+  );
+}