From 2a016f1ffd1af381ad340b9f98e967a7c66438e4 Mon Sep 17 00:00:00 2001
From: Jemma Stachelek <jemma.stachelek@gmail.com>
Date: Tue, 16 Apr 2024 10:21:50 -0600
Subject: [PATCH] pull initial vote data

---
 .gitignore                             |  3 ++-
 Makefile                               |  5 +++++
 hydropop/streamflow/selecting_gages.py | 23 ++++++++++++-----------
 3 files changed, 19 insertions(+), 12 deletions(-)

diff --git a/.gitignore b/.gitignore
index ca374a7..b2876db 100644
--- a/.gitignore
+++ b/.gitignore
@@ -154,4 +154,5 @@ results/*.shp
 results/*.tif
 results/*.csv
 results/*.shx
-data
\ No newline at end of file
+data
+hydropop/config.py
\ No newline at end of file
diff --git a/Makefile b/Makefile
index f79e3cf..91394df 100644
--- a/Makefile
+++ b/Makefile
@@ -10,3 +10,8 @@ clean:
 
 rabpro:
 	pip install --upgrade -e ../rabpro
+
+initial_data: data/initial_basins.gpkg data/initial_gages.gpkg
+
+data/initial_basins.gpkg data/initial_gages.gpkg: hydropop/streamflow/selecting_gages.py
+	python $<
diff --git a/hydropop/streamflow/selecting_gages.py b/hydropop/streamflow/selecting_gages.py
index a7af1ae..e30cc38 100644
--- a/hydropop/streamflow/selecting_gages.py
+++ b/hydropop/streamflow/selecting_gages.py
@@ -1,12 +1,13 @@
+import os
+import sys
 import geopandas as gpd
-from VotE import config
 
+sys.path.append(".")
+from hydropop import config
 config.vote_db()
 from VotE.streamflow import export_streamflow as es
-import os
-
 
-path_bounding_box = r"X:\Research\CIMMID\Data\Hydropop Layers\Finals\toronto_coarse\roi.gpkg"  # shapefile of ROI
+path_bounding_box = r"data/roi_small.gpkg"
 bb = gpd.read_file(path_bounding_box)
 gage_params = {
     "within": bb.geometry.values[0],
@@ -44,22 +45,22 @@
 # gages = es.get_gages(current_gages['id_gage'].values.tolist())
 
 gage_locs.to_file(
-    r"X:\Research\CIMMID\Data\Watersheds\Toronto\initial_gages.gpkg", driver="GPKG"
+    r"data/initial_gages.gpkg", driver="GPKG"
 )
 basins.to_file(
-    r"X:\Research\CIMMID\Data\Watersheds\Toronto\initial_basins.gpkg", driver="GPKG"
+    r"data/initial_basins.gpkg", driver="GPKG"
 )
 
 """ Download streamflow data """
+path_dir = "data/na_10k/streamflow/"
+os.makedirs(path_dir, exist_ok=True)
+
 for id_gage in basins["id_gage"].values:
     print(id_gage)
     df = es.get_streamflow_timeseries(
         [int(id_gage)], start_date="1981-01-01", expand=True, trim=True
     )
     df.drop("id_gage", inplace=True, axis=1)
-    df.sort_values(by="date", inplace=True)
-    path_out = os.path.join(
-        r"X:\Research\CIMMID\Data\Hydropop Layers\Finals\na_10k\streamflow",
-        str(id_gage) + ".csv",
-    )
+    df.sort_values(by="date", inplace=True)    
+    path_out = path_dir + str(id_gage) + ".csv"
     df.to_csv(path_out, index=False)