From 205881f536eff68ba7baffb720fe33c7e9962cfd Mon Sep 17 00:00:00 2001
From: yashgadhiya10 <ygadhiya@umd.edu>
Date: Thu, 29 Aug 2024 22:36:50 +0530
Subject: [PATCH] Created new Dataset North-Uganda-2017

---
 data/raw.dvc |  6 +++---
 datasets.py  | 32 ++++++++++++++++++++++++++++++++
 2 files changed, 35 insertions(+), 3 deletions(-)

diff --git a/data/raw.dvc b/data/raw.dvc
index bc5cdb43..f13f5629 100644
--- a/data/raw.dvc
+++ b/data/raw.dvc
@@ -1,6 +1,6 @@
 outs:
-- md5: 7ba2a5b698999a9bfa635252b5090ff7.dir
-  size: 446381456
-  nfiles: 406
+- md5: 0916e33f6eef6c80a87e319427005f5e.dir
+  size: 446720790
+  nfiles: 408
   path: raw
   hash: md5
diff --git a/datasets.py b/datasets.py
index 7e690524..920cbc46 100644
--- a/datasets.py
+++ b/datasets.py
@@ -506,6 +506,37 @@ def load_labels(self) -> pd.DataFrame:
         df[START], df[END] = date(2016, 1, 1), date(2017, 12, 31)
         df[SUBSET] = train_val_test_split(df.index, 0.3, 0.3)
         return df
+    
+class Uganda_NorthCEO2017(LabeledDataset):
+    def load_labels(self) -> pd.DataFrame:
+        raw_folder = raw_dir / "Uganda_North_2017"
+        df1 = pd.read_csv(
+            raw_folder
+            / "ceo-UNHCR-North-Uganda-Feb-2017---Feb-2018-(Set-1)-sample-data-2024-08-29.csv"
+        )
+        df2 = pd.read_csv(
+            raw_folder
+            / "ceo-UNHCR-North-Uganda-Feb-2017---Feb-2018-(Set-2)-sample-data-2024-08-29.csv"
+        )
+        df = pd.concat([df1, df2])
+
+        # Discard rows with no label
+        df = df[~df["Does this pixel contain active cropland?"].isna()].copy()
+        df[CLASS_PROB] = df["Does this pixel contain active cropland?"] == "Crop"
+        df[CLASS_PROB] = df[CLASS_PROB].astype(int)
+        df["num_labelers"] = 1
+        df = df.groupby([LON, LAT], as_index=False, sort=False).agg(
+            {
+                CLASS_PROB: "mean",
+                "num_labelers": "sum",
+                "plotid": join_unique,
+                "sampleid": join_unique,
+                "email": join_unique,
+            }
+        )
+        df[START], df[END] = date(2017, 1, 1), date(2018, 12, 31)
+        df[SUBSET] = train_val_test_split(df.index, 0.3, 0.3)
+        return df
 
 
 class Uganda_NorthCEO2021(LabeledDataset):
@@ -1538,6 +1569,7 @@ def load_labels(self) -> pd.DataFrame:
     TanzaniaCropArea2019(),
     FranceCropArea2020(),
     Uganda_NorthCEO2016(),
+    Uganda_NorthCEO2017(),
 ]
 
 if __name__ == "__main__":