Visualizeinit

datasets.py

@@ -137,8 +137,9 @@ def prepare_bosch(dataset_folder, nrows):
 
     os.system("kaggle competitions download -c bosch-production-line-performance -f " +
               filename + " -p " + dataset_folder)
-    X = pd.read_csv(local_url, index_col=0, compression='zip', dtype=np.float32,
-                    nrows=nrows)
+    X = pd.read_csv(local_url,compression='zip', dtype=np.float32)
+    X = X.set_index('Id')
+    X.index = X.index.astype('int64')
     y = X.iloc[:, -1].to_numpy(dtype=np.float32)
     X.drop(X.columns[-1], axis=1, inplace=True)
     X = X.to_numpy(dtype=np.float32)

visualization.py

@@ -0,0 +1,116 @@
+import os
+import sys
+import json
+import numpy as np
+import pandas as pd
+import argparse
+import matplotlib.pyplot as plt
+import csv
+
+def parse_args():
+    parser = argparse.ArgumentParser(
+        description="Visualize benchmarks against another version")
+    parser.add_argument("-d1", required=True, type=str,
+                        help="The first csv file directory with results")
+    parser.add_argument("-d2", required=True, type=str,
+                        help="The second csv file directory with results")
+    parser.add_argument("-metric", default="train_time", type=str,
+                        help=("The metric we want to visulaize"))
+    parser.add_argument("-title", default="graph", type=str,
+                        help=("The title of the graph"))
+    parser.add_argument("-output", default=sys.path[0] + "/results.png", type=str,
+                        help="Output json file with visualization")
+    args = parser.parse_args()
+    return args
+
+def autolabel(rects):
+    """Attach a text label above each bar in *rects*, displaying its height."""
+    for rect in rects:
+        height = rect.get_height()
+        ax.annotate('{}'.format(height),
+                    xy=(rect.get_x() + rect.get_width() / 2, height),
+                    xytext=(0, 3),  # 3 points vertical offset
+                    textcoords="offset points",
+                    ha='center', va='bottom')
+
+def algo_trimmer(algo):
+    print(f"algo: {algo}")
+    s = ""
+    for i in range(len(algo)):
+        if(algo[i] == '_'):
+            break
+        s+=algo[i]
+    return s
+
+def plot_error_bars(df, args):
+    gp = df.groupby("dataset")
+    means = gp.mean()
+    errors = gp.std()
+    if(args.metric == "train-time"):
+        fig, ax = plt.subplots()
+        pl = means.plot.bar(yerr=errors, ax=ax, capsize=10, rot=0, logy=True)
+        for idx, label in enumerate(list(means.index)): 
+            for acc in means.columns:
+                value = np.round(means.iloc[idx,0]/means.iloc[idx, 1],decimals=2)
+                ax.annotate(value,
+                            (idx, value),
+                            xytext=(0, 15), 
+                            textcoords='offset points')
+    else:
+        fig, ax = plt.subplots()
+        pl = means.plot.bar(yerr=errors, ax=ax, capsize=10, rot=0, figsize=(12, 12))
+        for p in ax.patches:
+            h = p.get_height()
+            x = p.get_x()+p.get_width()/2.
+            if h != 0:
+                ax.annotate("%g" % round(p.get_height(), 2), xy=(x,h), xytext=(0,4), rotation=90,
+                   textcoords="offset points", ha="center", va="bottom")
+        ax.legend(ncol=len(df.columns), loc="lower left", bbox_to_anchor=(0,1.02,1,0.08), 
+          borderaxespad=0, mode="expand")
+    plt.ylabel(f"{args.metric}")
+    plt.title(args.title) 
+    plt.savefig(args.output)
+
+def train_timer(args):
+    df = pd.DataFrame()
+    for idx, i in enumerate([args.d1, args.d2]):
+        temp_df = pd.read_csv(i)
+        temp_df = temp_df.drop(columns=["test_time", "AUC", "Accuracy", "F1","Precision",
+        "Recall","MeanAbsError","MeanSquaredError","MedianAbsError", "algorithm"])
+        if(idx == 0):
+            temp_df = temp_df.rename(columns={"train_time": f"{args.d1}_train_time"})
+            df = df.append(temp_df)
+        else:
+            temp_df = temp_df.rename(columns={"train_time": f"{args.d2}_train_time"})
+            df = df.merge(temp_df, on="dataset")
+    return df
+
+def accuracy_timer(args):
+    df = pd.DataFrame()
+    for idx, i in enumerate([args.d1, args.d2]):
+        temp_df = pd.read_csv(i)
+        temp_df = temp_df.drop(columns=["test_time", "AUC", "train_time", "F1","Precision",
+        "Recall","MeanAbsError","MeanSquaredError","MedianAbsError", "algorithm"])
+        if(idx == 0):
+            temp_df = temp_df[temp_df["Accuracy"] != "-na-"]
+            temp_df = temp_df.rename(columns={"Accuracy": f"{args.d1}_Accuracy"})
+            df = df.append(temp_df)
+        else:
+            temp_df = temp_df[temp_df["Accuracy"] != "-na-"]
+            temp_df = temp_df.rename(columns={"Accuracy": f"{args.d2}_Accuracy"})
+            df = df.merge(temp_df, on="dataset")
+    df = df.astype({f"{args.d1}_Accuracy": 'float32', f"{args.d2}_Accuracy": 'float32'})
+    return df
+
+def main():
+    args = parse_args()
+    df = pd.DataFrame()
+    if(args.metric == 'train-time'):
+        df = train_timer(args)
+    else:
+        df = accuracy_timer(args)
+    plot_error_bars(df, args)
+
+
+if __name__ == '__main__':
+    main()