Yo, welcome to the PyQuery UI. This is where you can click buttons and make things happen. It's built on Streamlit, so it's snappy and looks fire.
When you launch the app (pyquery ui), you'll see a clean interface.
This is where you load your data.
- File Loader: Drag and drop your CSVs, Excels, or Parquets here. We support multi-file uploads, so don't be shy.
- Dataset List: See what's currently loaded in memory. Click to switch context.
This is the default tab.
- Data Preview: See the first 1000 rows of your data.
- Transforms via Column Headers: Click a column header to filter, sort, or rename.
- Recipe Editor: Add steps manually if you're feeling spicy.
- Export: When you're done cooking, download your data as Parquet, CSV, or Excel.
Get intimate with your data.
- Univariate: Look at one column at a time. Histograms, box plots, the works.
- Bivariate: Compare two columns. Scatter plots, correlation heatmaps.
- Contrast: See how your data differs across categories.
If you speak SQL, this tab is for you.
- Write standard SQL queries against your loaded dataframes.
- We use DuckDB/Polars under the hood, so it's wicked fast.
- Your dataframes are available as tables (e.g.,
SELECT * FROM my_dataset).
Get a full report card on your dataset.
- Missing values, unique counts, duplicates.
- It's like a physical exam for your data.
- Load: Drop files in the sidebar.
- Cook: Use the Recipe tab to clean and transform.
- Check: pop over to EDA/Profiling to make sure you didn't mess up.
- Export: Download the result and flex on your colleagues.