Refactor the code

EIA_Energy_Data_Analysis.ipynb

@@ -10,8 +10,10 @@
     "showInput": false
    },
    "source": [
+    "`Copyright (c) Meta Platforms, Inc. and affiliates. This source code is licensed under the CC-BY-NC license found in the LICENSE file in the root directory of this source tree.`\n",
+    "\n",
     "## This notebook downloads and analyzes EIA's bulk U.S. Electric System Operating Data \n",
-    "### [https://www.eia.gov/opendata/bulkfiles.php]\n"
+    "For description of the data, see: https://www.eia.gov/opendata/bulkfiles.php\n"
    ]
   },
   {

src/cas.py

@@ -0,0 +1,80 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# This source code is licensed under the CC-BY-NC license found in the
+# LICENSE file in the root directory of this source tree.
+
+import pandas as pd
+
+# Carbon Aware Scheduling Algorithm
+# takes a dataframe that contains renewable and dc power, dc_all
+# applies cas within the flexible_workload_ratio, and max_capacity constraints
+# returns the carbon balanced version of the input dataframe, balanced_df
+def cas(df_all, flexible_workload_ratio, max_capacity):
+    # work on 24 hour basis
+    # sort the df in terms of ascending renewable en
+    # take flexible_workload_ratio from the highest carbon intensity hours
+    # to lowest ones if there is not enough renewables until max_capacity is hit
+    balanced_df = []
+    for i in range(0, df_all.shape[0], 24):
+        sorted_df = df_all[i : i + 24].sort_values(
+            by=["tot_renewable", "avg_dc_power_mw"]
+        )
+        start = 0
+        end = 23
+        if sorted_df.shape[0] < 23:
+            break
+        work_to_move = 0
+        while start < end:
+            renewable_surplus = (
+                sorted_df["tot_renewable"].iloc[end]
+                - sorted_df["avg_dc_power_mw"].iloc[end]
+            )
+            renewable_gap = (
+                sorted_df["avg_dc_power_mw"].iloc[start]
+                - sorted_df["tot_renewable"].iloc[start]
+            )
+            available_space = min(
+                renewable_surplus,
+                (max_capacity - sorted_df["avg_dc_power_mw"].iloc[end]),
+            )
+            if renewable_surplus <= 0:
+                end = end - 1
+                continue
+            if renewable_gap <= 0:
+                start = start + 1
+                continue
+            if work_to_move <= 0 and renewable_gap > 0:
+                work_to_move = min(
+                    renewable_gap,
+                    (
+                        flexible_workload_ratio
+                        / 100
+                        * sorted_df["avg_dc_power_mw"].iloc[start]
+                    ),
+                )
+
+            if available_space > work_to_move:
+                sorted_df["avg_dc_power_mw"].iloc[end] = (
+                    sorted_df["avg_dc_power_mw"].iloc[end]
+                    + work_to_move
+                )
+                sorted_df["avg_dc_power_mw"].iloc[start] = (
+                    sorted_df["avg_dc_power_mw"].iloc[start]
+                    - work_to_move
+                )
+                start = start + 1
+                work_to_move = 0
+            else:
+                sorted_df["avg_dc_power_mw"].iloc[end] = (
+                    sorted_df["avg_dc_power_mw"].iloc[end]
+                    + available_space
+                )
+                sorted_df["avg_dc_power_mw"].iloc[start] = (
+                    sorted_df["avg_dc_power_mw"].iloc[start]
+                    - available_space
+                )
+                work_to_move = work_to_move - available_space
+                end = end - 1
+        balanced_df.append(sorted_df)
+
+    final_balanced_df = pd.concat(balanced_df).sort_values(by=["index"])
+    return final_balanced_df

src/download_and_process.py

@@ -1,3 +1,7 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# This source code is licensed under the CC-BY-NC license found in the
+# LICENSE file in the root directory of this source tree.
+
 import wget
 import zipfile
 import pandas as pd