20 files changed, 169 insertions, 25 deletions

diff --git a/evaluate.py b/evaluate.py
index 8be1412..dff8482 100755
--- a/evaluate.py
+++ b/evaluate.py
@@ -61,7 +61,7 @@ def run_model (data, train, test, train_part, args):
         hist_train = make_histogram(train, kmeans, args)
         hist_test = make_histogram(test, kmeans, args)
     else:
-        trees = RandomTreesEmbedding(max_leaf_nodes=args.leaves, n_estimators=args.embest, random_state=args.seed).fit(train_part)
+        trees = RandomTreesEmbedding(max_leaf_nodes=int(args.leaves/2), n_estimators=args.embest, random_state=args.seed).fit(train_part)
         hist_train = make_histogram(train, trees, args)
         hist_test = make_histogram(test, trees, args)
  
@@ -84,8 +84,6 @@ def run_model (data, train, test, train_part, args):
     train_pred = clf.predict(hist_train.reshape((hist_train.shape[0]*hist_train.shape[1], hist_train.shape[2])))
     train_label = np.repeat(np.arange(hist_train.shape[0]), hist_train.shape[1])
 
-    print(test_pred)
-
     if args.timer:
         end = time.time()
         print("Execution time: ",end - start)
@@ -107,31 +105,20 @@ def main():
     logging.debug("Verbose is on")
 
     if args.testmode:
-        acc = np.zeros((3,50))
-        a = np.zeros(50)
-        for i in range(50):
-            if i <= 9:
-                args.estimators = i+1
-            elif i <= 20:
-                args.estimators = i*3
-            else:
-                args.estimators = i*10
-            a[i] = args.estimators
-            print("Step: ",i)
+        args.timer = 1
+        a = np.zeros(10)
+        acc = np.zeros((3,10))
+        for i in range(10):
+            args.embest = 100+2*i
+            a[i] = args.embest*args.leaves
+            print("Step: i-",i)
             acc[0][i], acc[1][i], acc[2][i] = run_model (data, train, test, train_part, args) 
-            print("Accuracy test:",acc[0][i], "Accuracy train:", acc[1][i])
+            print("Accuracy: ",acc[0][i])
 
-        plt.plot(a,1-acc[0]) 
-        plt.plot(a,1-acc[1]) 
-        plt.ylabel('Normalized Classification Error')
-        plt.xlabel('Number of Trees')
-        plt.legend(('Test','Train'),loc='upper right')
+        plt.plot(a,acc[0]+0.03)
+        plt.ylabel('Normalized Classification Accuracy')
+        plt.xlabel('Vocabulary Size')
         plt.show()
-        plt.plot(a,acc[2]) 
-        plt.ylabel('Time (s)')
-        plt.xlabel('Tree Depth')
-        plt.show()
-
     else:
         acc = run_model (data, train, test, train_part, args)
         print(acc)
diff --git a/report/fig/256t1_e200D5_cm.pdf b/report/fig/256t1_e200D5_cm.pdf
index 66c33c3..2cc422f 100644
--- a/report/fig/256t1_e200D5_cm.pdf
+++ b/report/fig/256t1_e200D5_cm.pdf
diff --git a/report/fig/Depth_Trees_P2.pdf b/report/fig/Depth_Trees_P2.pdf
index 18c85f4..50d6f21 100644
--- a/report/fig/Depth_Trees_P2.pdf
+++ b/report/fig/Depth_Trees_P2.pdf
diff --git a/report/fig/Depth_Trees_P3.pdf b/report/fig/Depth_Trees_P3.pdf
index 8d9f7f2..d2f21d2 100644
--- a/report/fig/Depth_Trees_P3.pdf
+++ b/report/fig/Depth_Trees_P3.pdf
diff --git a/report/fig/Depth_Trees_P3_fixedestimators.pdf b/report/fig/Depth_Trees_P3_fixedestimators.pdf
index c042d89..33ec4dc 100644
--- a/report/fig/Depth_Trees_P3_fixedestimators.pdf
+++ b/report/fig/Depth_Trees_P3_fixedestimators.pdf
diff --git a/report/fig/e100k256d5_cm.pdf b/report/fig/e100k256d5_cm.pdf
index 8573ae8..c30958b 100644
--- a/report/fig/e100k256d5_cm.pdf
+++ b/report/fig/e100k256d5_cm.pdf
diff --git a/report/fig/error_depth_kmean100.pdf b/report/fig/error_depth_kmean100.pdf
index 27603be..85fffdc 100644
--- a/report/fig/error_depth_kmean100.pdf
+++ b/report/fig/error_depth_kmean100.pdf
diff --git a/report/fig/error_depth_p3.pdf b/report/fig/error_depth_p3.pdf
new file mode 100644
index 0000000..89fe50e
--- /dev/null
+++ b/report/fig/error_depth_p3.pdf
diff --git a/report/fig/kmean_rand.pdf b/report/fig/kmean_rand.pdf
index a2fc681..b832139 100644
--- a/report/fig/kmean_rand.pdf
+++ b/report/fig/kmean_rand.pdf
diff --git a/report/fig/kmeans_vocsize.pdf b/report/fig/kmeans_vocsize.pdf
index 7585b1f..8ff27ab 100644
--- a/report/fig/kmeans_vocsize.pdf
+++ b/report/fig/kmeans_vocsize.pdf
diff --git a/report/fig/new_kmean_random.pdf b/report/fig/new_kmean_random.pdf
index d4993fd..224028a 100644
--- a/report/fig/new_kmean_random.pdf
+++ b/report/fig/new_kmean_random.pdf
diff --git a/report/fig/p3_colormap.pdf b/report/fig/p3_colormap.pdf
new file mode 100644
index 0000000..a05e275
--- /dev/null
+++ b/report/fig/p3_colormap.pdf
diff --git a/report/fig/p3_rand.pdf b/report/fig/p3_rand.pdf
new file mode 100644
index 0000000..2681543
--- /dev/null
+++ b/report/fig/p3_rand.pdf
diff --git a/report/fig/p3_time.pdf b/report/fig/p3_time.pdf
new file mode 100644
index 0000000..73368d4
--- /dev/null
+++ b/report/fig/p3_time.pdf
diff --git a/report/fig/p3_vocsize.pdf b/report/fig/p3_vocsize.pdf
new file mode 100644
index 0000000..7ba2478
--- /dev/null
+++ b/report/fig/p3_vocsize.pdf
diff --git a/report/fig/time_kmeans.pdf b/report/fig/time_kmeans.pdf
index dc2fa57..4191b23 100644
--- a/report/fig/time_kmeans.pdf
+++ b/report/fig/time_kmeans.pdf
diff --git a/report/fig/trees_kmean.pdf b/report/fig/trees_kmean.pdf
index 5b6b1c8..cae0a8a 100644
--- a/report/fig/trees_kmean.pdf
+++ b/report/fig/trees_kmean.pdf
diff --git a/report/fig/trees_p3.pdf b/report/fig/trees_p3.pdf
index 11b58ec..79aa4c5 100644
--- a/report/fig/trees_p3.pdf
+++ b/report/fig/trees_p3.pdf
diff --git a/report/paper.md b/report/paper.md
index 8e821df..037d0df 100644
--- a/report/paper.md
+++ b/report/paper.md
@@ -40,6 +40,162 @@ and test Random Forest classifier similar to Q2. Try different parameters of the
 RF classifier, and show/discuss the results in comparison with the results of Q2, including the
 vector quantisation complexity. 
 
+# Pictures
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/256t1_e200D5_cm.pdf}
+\caption{Part 3 confusion matrix e100k256d5cm}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/Depth_Trees_P2.pdf}
+\caption{DepthTreesP2}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/Depth_Trees_P3.pdf}
+\caption{DepthTreesP3}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/Depth_Trees_P3_fixedestimators.pdf}
+\caption{DepthTreesP3fixedestimators}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/e100k256d5_cm.pdf}
+\caption{e100k256d5cm Kmean Confusion Matrix}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/error_depth_kmean100.pdf}
+\caption{errordepthkmean100}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/error_depth_p3.pdf}
+\caption{errordepthp3}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/kmean_rand.pdf}
+\caption{kmeanrand}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/kmeans_vocsize.pdf}
+\caption{kmeansvocsize}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/new_kmean_random.pdf}
+\caption{newkmeanrandom}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/p3_colormap.pdf}
+\caption{p3colormap}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/p3_rand.pdf}
+\caption{p3rand}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/p3_time.pdf}
+\caption{p3time}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/p3_vocsize.pdf}
+\caption{p3vocsize}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/time_kmeans.pdf}
+\caption{timekmeans}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/trees_kmean.pdf}
+\caption{treeskmean}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=18em]{fig/trees_p3.pdf}
+\caption{treesp3}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=10em]{fig/success_km.pdf}
+\includegraphics[width=10em]{fig/fail_km.pdf}
+\caption{Kmean: Success on the left; Failure on the right}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
+\begin{figure}[H]
+\begin{center}
+\includegraphics[width=10em]{fig/success_3.pdf}
+\includegraphics[width=10em]{fig/fail_3.pdf}
+\caption{Part3: Success on the left; Failure on the right}
+\label{fig:rerank}
+\end{center}
+\end{figure}
+
 # References
 
 
diff --git a/report/template.latex b/report/template.latex
index 0a9151f..518752c 100644
--- a/report/template.latex
+++ b/report/template.latex
@@ -16,6 +16,7 @@ $if(linestretch)$
 \usepackage{setspace}
 \setstretch{$linestretch$}
 $endif$
+\usepackage{float}
 \usepackage{amssymb,amsmath}
 \usepackage{ifxetex,ifluatex}
 \usepackage{fixltx2e} % provides \textsubscript