should be it

external/duckdb/benchmark/trainbenchmark/README.md

@@ -0,0 +1,27 @@
+# Train Benchmark queries
+
+For the benchmark specification see the paper [The Train Benchmark: cross-technology performance evaluation of continuous model queries](https://link.springer.com/10.1007/s10270-016-0571-8) (Softw. Syst. Model. 2018)
+
+## Usage
+
+### Loading the data
+
+The test data is from the benchmark's "Repair" scenario with data set size (scale factor) 1.
+
+```bash
+cat schema.sql | duckdb tb.duckdb
+TB_DATA_DIR=sf1
+sed -i.bkp "s|\\N||g" ${TB_DATA_DIR}/Route.csv
+sed "s/PATHVAR/${TB_DATA_DIR}/g" tb-load.sql | duckdb tb.duckdb
+```
+
+### Queries
+
+```bash
+cat queries/connectedsegments.sql | duckdb tb.duckdb
+cat queries/routesensor.sql | duckdb tb.duckdb
+cat queries/poslength.sql | duckdb tb.duckdb
+cat queries/semaphoreneighbor.sql | duckdb tb.duckdb
+cat queries/switchmonitored.sql | duckdb tb.duckdb
+cat queries/switchset.sql | duckdb tb.duckdb
+```