example analysis for #57

scripts/plots/ofe_results.py

@@ -0,0 +1,42 @@
+"""Some exploritory OFE based results"""
+from __future__ import print_function
+import pandas as pd
+import numpy as np
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+
+def rotclass(val):
+    """Convert this rotation string into a classification"""
+    # if we have seven years of either corn and soy
+    if val.count("B") + val.count("C") > 6:
+        return "A"
+    return "N"
+
+
+def main():
+    """Go Main Go"""
+    dfin = pd.read_csv('results.csv')
+    # Filter data to only include slopes < 0.5 and delivery values < 20 T/a/yr
+    df = dfin[(dfin['slope[1]'] < 0.5) &
+              (dfin['delivery[t/a/yr]'] < 20)].copy()
+    # classify rotation
+    df['rotclass'] = df['CropRotationString'].apply(rotclass)
+    bins = [0, 0.1, 0.2, 0.3, 0.4, 0.5]
+    df['slpclass'] = np.searchsorted(bins, df['slope[1]'].values)
+    ax = sns.factorplot(kind='box',        # Boxplot
+                         y='delivery[t/a/yr]',
+                         x='slpclass',
+                         hue='rotclass',
+                         data=df,
+                         size=8,
+                         aspect=1.5,
+                         legend_out=False)
+    ax.set_xticklabels(["%s-%s" % (bins[i], bins[i+1])
+                        for i in range(len(bins)-1)])
+    # sns.jointplot(x='slope[1]', y='delivery[t/a/yr]', data=df, kind='reg')
+    plt.savefig('test.png')
+
+
+if __name__ == '__main__':
+    main()