Add tufts-weather benchmark

tuft-weather/.gitignore

@@ -0,0 +1,2 @@
+*.png
+*.txt

tuft-weather/clean.sh

@@ -0,0 +1,2 @@
+rm *.txt
+rm *.png

tuft-weather/run.sh

@@ -0,0 +1,30 @@
+#!/usr/bin/env bash
+
+input="input"
+
+cat "$input" \
+    | grep -v "\-99" \
+    | sort -n \
+    | awk '{ printf "%02d-%02d %s %s\n", $1, $2, $3, $4 }' > formatted.txt
+
+input="formatted.txt"
+
+# Get max and min per month date across all years
+cat "$input" \
+    | awk '{
+        key = sprintf("%s", $1);
+        if (!(key in max) || $3 > max[key]) max[key] = $3;
+        if (!(key in min) || $3 < min[key]) min[key] = $3;
+    } 
+    END {
+        for (key in max) {
+            printf "%s %s %s\n", key, max[key], min[key];
+        }
+    }' \
+    | sort -n > max_min.txt
+
+for year in $(seq 1995 2013); do
+    join -1 1 -2 1 -o "0 1.2 1.3 2.2" -e "NA" max_min.txt <(grep "$year" "$input" | cut -d' ' -f 1,3) \
+    | grep -v "NA" \
+    | ./scripts/plot.py "$year" > "$year.png"
+done

tuft-weather/scripts/plot.py

@@ -0,0 +1,46 @@
+#!/usr/bin/env python3
+
+import datetime
+import io
+import matplotlib.pyplot as plt
+import numpy as np
+import sys
+
+# Read data from stdin
+data = sys.stdin
+year = sys.argv[1]
+
+to_dt = lambda x: datetime.datetime.strptime(f"{year}-{x}", "%Y-%m-%d")
+
+data = [ line.strip().replace("\r", "") for line in data ]
+data = [ (to_dt(date), float(_min), float(_max), float(temp)) for date, _min, _max, temp in [ line.split() for line in data if line ] ]
+data = { key: [ x for x in values ] for key, values in zip(["date", "min", "max", "temp"], zip(*data)) }
+
+dates = data["date"]
+historic_max = data["max"]
+historic_min = data["min"]
+temps = data["temp"]
+
+# Plotting
+plt.figure(figsize=(10, 5))
+
+# Historic ranges (shaded areas)
+plt.fill_between(dates, historic_min, historic_max, color="tan", alpha=0.3, label='Normal range')
+
+plt.plot(dates, temps, color="black", label=f"{year} temperature")
+
+# Labels and titles
+plt.xlabel(f"Days in {year}")
+plt.ylabel("Temperature (°F)")
+plt.title(f"Temperature Records for {year}")
+
+# Legend
+plt.legend(loc="upper right")
+
+# Show the plot
+plt.grid(True)
+
+buf = io.BytesIO()
+plt.savefig(buf, format='png')
+plt.close()
+sys.stdout.buffer.write(buf.getvalue())