From 0fa3c29253bb89759b38b919df20373df1424f14 Mon Sep 17 00:00:00 2001
From: Khalim Conn-Kowlessar <kconnkowlessar@gmail.com>
Date: Fri, 16 Jun 2023 17:29:48 +0100
Subject: [PATCH] Tidy up app script removing geopandas, geopy

---
 model_data/app.py           | 60 +++++++++++++------------------------
 model_data/requirements.txt |  3 +-
 2 files changed, 22 insertions(+), 41 deletions(-)

diff --git a/model_data/app.py b/model_data/app.py
index f03d283d..23fc49fe 100644
--- a/model_data/app.py
+++ b/model_data/app.py
@@ -128,22 +128,34 @@ def handler():
     table = DBF(borehole_file)
     borehole_data = [x for x in tqdm(table, total=len(table))]
 
+    # There are ~1.4 million entries in this dataset and so we firstly want to reduce the number of
+    # entries in here if possible before we produce any form of comparison between our properties, to infer
+    # the distance from the property to the nearest borehole
+
     # Let's take a sample
     borehole_sample = borehole_data[:1000]
     df = pd.DataFrame(borehole_sample)
 
     # for each property, find the nearest borehole
-    # Data dictionary:
-    # QS
-    # NUMB
-    # BSUFF
+
+    # Data dictionary: This description is based on the information presented in the following
+    # Geological articles:
+    # https://nora.nerc.ac.uk/id/eprint/509366/1/IR04115.pdf
+    # https://shop.bgs.ac.uk/Resources/Shop/doc/info/Borehole_Abbreviations.pdf?_ga=2.246788941.895115819.1686912089
+    # -542796874.1686912089
+    # https://core.ac.uk/download/63732.pdf
+    #
+    #
+    # QS - Borehole identifier information
+    # NUMB - Borehole identifier information
+    # BSUFF - Borehole identifier information
     # REGNO
-    # RT
+    # RT - Borehole identifier information
     # GRID_REFER
-    # EASTING
-    # NORTHING
-    # X
-    # Y
+    # EASTING - British National Grid coordinates
+    # NORTHING - British National Grid coordinates
+    # X - British National Grid coordinates - same as EASTING but has a float typing
+    # Y - British National Grid coordinates - same as NORTHING but has a float typing
     # CONFIDENTI
     # STRTHEIGHT
     # NAME
@@ -153,33 +165,3 @@ def handler():
     # DATE_K_TYP
     # DATE_ENTER
     # AGS_LOG_UR
-
-    import geopandas as gpd
-
-    borehole_shape_file = os.path.abspath(os.path.dirname(__file__)) + "/model_data/local_data/borehole/borehole.shp"
-    boreholes = gpd.read_file(borehole_shape_file)
-
-    from geopy.geocoders import Nominatim
-    from geopy.extra.rate_limiter import RateLimiter
-
-    geolocator = Nominatim(user_agent="myGeocoder")
-
-    # The rate limiter will ensure you're not making too many requests too quickly
-    geocode = RateLimiter(geolocator.geocode, min_delay_seconds=0.5)
-
-    for address_config in input_data:
-        location = geocode(" ".join([address_config["address1"], address_config["postcode"]]))
-        geocode("28 distillery wharf, regtta lane, hammersmith, london, w6 9bf")
-
-        print((location.latitude, location.longitude))
-
-    from geopy.geocoders import Nominatim
-    from geopy.extra.rate_limiter import RateLimiter
-
-    geolocator = Nominatim(user_agent="myGeocoder")
-
-    geocode = RateLimiter(geolocator.geocode, min_delay_seconds=1)
-
-    # Full address
-    location = geocode(", ".join([p.data["address"], p.data["postcode"]]))
-    print((location.latitude, location.longitude))
diff --git a/model_data/requirements.txt b/model_data/requirements.txt
index 9888d372..a21cdf6e 100644
--- a/model_data/requirements.txt
+++ b/model_data/requirements.txt
@@ -10,5 +10,4 @@ pytest-mock
 fuzzywuzzy
 python-Levenshtein
 dbfread
-geopandas
-geopy
\ No newline at end of file
+pyproj
\ No newline at end of file