Let user set number of rows in simulated data (#183)

* add selector for row count

* add to test; fix fragile test

dp_wizard/app/analysis_panel.py

@@ -27,6 +27,16 @@ def analysis_ui():
             [],
         ),
         ui.output_ui("columns_ui"),
+        ui.markdown(
+            "What is the approximate number of rows in the dataset? "
+            "This number is only used for the simulation and not the final calculation."
+        ),
+        ui.input_select(
+            "row_count",
+            "Estimated Rows",
+            choices=["100", "1000", "10000"],
+            selected="100",
+        ),
         ui.markdown(
             "What is your privacy budget for this release? "
             "Values above 1 will add less noise to the data, "
@@ -108,6 +118,7 @@ def columns_ui():
                 name=column_ids_to_names[column_id],
                 contributions=contributions(),
                 epsilon=epsilon(),
+                row_count=int(input.row_count()),
                 lower_bounds=lower_bounds,
                 upper_bounds=upper_bounds,
                 bin_counts=bin_counts,
@@ -135,7 +146,7 @@ def columns_ui():
                 (
                     ui.layout_columns(
                         [],
-                        [ui.markdown(note_md)],
+                        ui.markdown(note_md),
                         col_widths=col_widths,  # type: ignore
                     )
                     if column_ids

dp_wizard/app/components/column_module.py

@@ -54,6 +54,7 @@ def column_server(
     name: str,
     contributions: int,
     epsilon: float,
+    row_count: int,
     lower_bounds: reactive.Value[dict[str, float]],
     upper_bounds: reactive.Value[dict[str, float]],
     bin_counts: reactive.Value[dict[str, int]],
@@ -166,6 +167,7 @@ def column_plot():
             # Exit early to avoid divide-by-zero.
             return None
         accuracy, histogram = make_accuracy_histogram(
+            row_count=row_count,
             lower=lower_x,
             upper=upper_x,
             bin_count=bin_count,

dp_wizard/utils/dp_helper.py

@@ -13,6 +13,7 @@
 
 
 def make_accuracy_histogram(
+    row_count: int,
     lower: float,
     upper: float,
     bin_count: int,
@@ -22,7 +23,12 @@ def make_accuracy_histogram(
     """
     Creates fake data between lower and upper, and then returns a DP histogram from it.
     >>> accuracy, histogram = make_accuracy_histogram(
-    ...     lower=0, upper=10, bin_count=5, contributions=1, weighted_epsilon=1)
+    ...     row_count=100,
+    ...     lower=0, upper=10,
+    ...     bin_count=5,
+    ...     contributions=1,
+    ...     weighted_epsilon=1
+    ... )
     >>> accuracy
     3.37...
     >>> histogram
@@ -42,7 +48,6 @@ def make_accuracy_histogram(
     # Mock data only depends on lower and upper bounds, so it could be cached,
     # but I'd guess this is dominated by the DP operations,
     # so not worth optimizing.
-    row_count = 100
     df = mock_data({"value": ColumnDef(lower, upper)}, row_count=row_count)
 
     # TODO: When this is stable, merge it to templates, so we can be

tests/test_app.py

@@ -85,6 +85,10 @@ def expect_no_error():
     download_results_button = page.get_by_role("button", name="Download results")
     assert download_results_button.is_disabled()
 
+    # Currently the only change when the estimated rows changes is the plot,
+    # but we could have the confidence interval in the text...
+    page.get_by_label("Estimated Rows").select_option("1000")
+
     # Set column details:
     page.get_by_label("grade").check()
     expect_visible(simulation)
@@ -103,7 +107,7 @@ def expect_no_error():
     assert page.get_by_label("Upper").input_value() == new_value
     # Add a second column:
     page.get_by_label("blank").check()
-    expect_visible("Weight")
+    expect(page.get_by_text("Weight")).to_have_count(2)
     # TODO: Setting more inputs without checking for updates
     # causes recalculations to pile up, and these cause timeouts on CI:
     # It is still rerendering the graph after hitting "Download results".