Add data pipeline example for efficient weather data handling

Author: Gnananjali

examples/data_pipeline/README.md

@@ -0,0 +1,24 @@
+# Data Pipeline Example for Weather Routing Tool
+
+This example demonstrates how to efficiently handle and process weather datasets using scalable techniques.
+
+## Overview
+
+Working with large weather datasets can be memory-intensive and slow when using traditional loading methods. This example shows how to improve performance and scalability by using:
+
+- Chunked data loading with Dask
+- Subsetting and interpolation
+- Efficient storage formats (NetCDF and Zarr)
+- Basic performance comparison
+
+## Features
+
+- Load dataset using chunking (`xarray + Dask`)
+- Subset a specific geographic region
+- Interpolate data at a given location
+- Save processed data in:
+  - NetCDF format
+  - Zarr format
+- Compare execution time for different storage methods
+
+## File Structure

examples/data_pipeline/analysis.py

@@ -0,0 +1,81 @@
+import xarray as xr
+import time
+import os
+
+# -------------------------------
+# LOAD DATASET WITH CHUNKING
+# -------------------------------
+print("\n--- Loading Dataset ---")
+ds = xr.open_dataset("weather.nc", chunks={"time": 10})
+
+print(ds)
+
+# -------------------------------
+# SUBSETTING REGION
+# -------------------------------
+ds = ds.sel(
+    latitude=slice(10, 20),
+    longitude=slice(80, 90)
+)
+
+print("\n--- After Subsetting ---")
+print(ds)
+
+# -------------------------------
+# INTERPOLATION
+# -------------------------------
+ds = ds.interp(latitude=15, longitude=85)
+
+print("\n--- After Interpolation ---")
+print(ds)
+
+# -------------------------------
+# SAVE AS NETCDF
+# -------------------------------
+print("\n--- Saving as NetCDF ---")
+start = time.time()
+
+ds.to_netcdf("output.nc")
+
+end = time.time()
+netcdf_time = end - start
+
+print("NetCDF Save Time:", netcdf_time, "seconds")
+
+# -------------------------------
+# SAVE AS ZARR
+# -------------------------------
+print("\n--- Saving as Zarr ---")
+start = time.time()
+
+ds.to_zarr("output.zarr", mode="w")
+
+end = time.time()
+zarr_time = end - start
+
+print("Zarr Save Time:", zarr_time, "seconds")
+
+# -------------------------------
+# FILE SIZE (NetCDF)
+# -------------------------------
+if os.path.exists("output.nc"):
+    nc_size = os.path.getsize("output.nc") / (1024 * 1024)
+    print("\nNetCDF File Size:", round(nc_size, 2), "MB")
+
+# -------------------------------
+# VARIABLES INFO
+# -------------------------------
+print("\n--- Dataset Variables ---")
+print(list(ds.data_vars))
+
+
+print("\n--- Performance Summary ---")
+print("NetCDF Time:", netcdf_time, "seconds")
+print("Zarr Time:", zarr_time, "seconds")
+
+if zarr_time < netcdf_time:
+    print("Zarr is faster for this dataset")
+else:
+    print("NetCDF is faster for this dataset")
+
+print("\n✔ Data pipeline example completed successfully")

examples/data_pipeline/output.nc

@@ -0,0 +1,46 @@
+{
+  "shape": [
+    25
+  ],
+  "data_type": "float64",
+  "chunk_grid": {
+    "name": "regular",
+    "configuration": {
+      "chunk_shape": [
+        10
+      ]
+    }
+  },
+  "chunk_key_encoding": {
+    "name": "default",
+    "configuration": {
+      "separator": "/"
+    }
+  },
+  "fill_value": "NaN",
+  "codecs": [
+    {
+      "name": "bytes",
+      "configuration": {
+        "endian": "little"
+      }
+    },
+    {
+      "name": "zstd",
+      "configuration": {
+        "level": 0,
+        "checksum": false
+      }
+    }
+  ],
+  "attributes": {
+    "coordinates": "latitude longitude",
+    "_FillValue": "AAAAAAAA+H8="
+  },
+  "dimension_names": [
+    "time"
+  ],
+  "zarr_format": 3,
+  "node_type": "array",
+  "storage_transformers": []
+}