Skip to content
Kharita Challenges

Kinds of problem and what you hand in

Nine ways a problem is answered, from one SQL query to a diagram you draw or a review you make, and which of them you can run.

Updated

The nine kinds

KindWhat you hand inBefore you submit
SQLOne SELECT query on the PostGIS tables the brief lists, with spatial output named geomRun shows the rows and the query plan
PythonA script that assigns its output to resultRun runs it in your browser
GeoJSONThe features you derived — pasted, or uploaded from GeoJSON, a zipped Shapefile, KML/KMZ, GPX or CSV—
ValuesNamed answers in a form, worked out in any tool you like—
SchemaPostgreSQL/PostGIS DDLRun builds a temporary schema and shows what it holds
DesignOne option for each decision, with a sentence on why — the choice is graded, the reasoning never is—
Prepared datasetA GeoPackage, zipped Shapefile, GeoJSON or CSV of up to 20 MBRead shows what it was read as
DiagramA system's components, an analysis workflow's steps, a deployment's machines or a geodatabase's classes, placed on a canvas and joined with arrows — see Draw a diagramList shows the same diagram as lists you can edit by keyboard
ReviewEvery feature of a delivered dataset that breaks a rule of its specification, flagged with the rule it breaks — see Review a deliveryYour flags lists them rule by rule

The CRS box

When a requirement checks the coordinate reference system, a Result CRS EPSG box appears beside the editor: say which system your answer is in.

Python's contract

Each dataset is at /data/<dataset id>.geojson, and what you assign to result is graded: a GeoDataFrame in any CRS (it is moved to WGS 84), a DataFrame, a dict of named values, or GeoJSON. Run executes in your browser; Submit runs the script again on the server and grades what result holds.

Still stuck? Write to Support