diff --git a/.travis.yml b/.travis.yml
index 53574a9..8a39511 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -5,9 +5,12 @@ notifications:
   email: false
 before_install:
   - sudo add-apt-repository ppa:staticfloat/julia-deps -y
-  - sudo add-apt-repository ppa:staticfloat/julianightlies -y
+  - sudo add-apt-repository ppa:staticfloat/juliareleases -y
   - sudo apt-get update -qq -y
   - sudo apt-get install libpcre3-dev julia -y
 script:
   - julia -e 'Pkg.init(); run(`ln -s $(pwd()) $(Pkg.dir("LinearLeastSquares"))`); Pkg.pin("LinearLeastSquares"); Pkg.resolve()'
   - julia -e 'using LinearLeastSquares; @assert isdefined(:LinearLeastSquares); @assert typeof(LinearLeastSquares) === Module'
+  - julia --code-coverage test/runtests.jl
+after_success:
+  - julia -e 'cd(Pkg.dir("LinearLeastSquares")); Pkg.add("Coverage"); using Coverage; Coveralls.submit(Coveralls.process_folder())'
diff --git a/REQUIRE b/REQUIRE
index 93db550..e3edc9e 100644
--- a/REQUIRE
+++ b/REQUIRE
@@ -1,2 +1,2 @@
-julia 0.2+
+julia 0.3
 
diff --git a/examples/regression.jl b/examples/regression.jl
deleted file mode 100644
index aeb26a1..0000000
--- a/examples/regression.jl
+++ /dev/null
@@ -1,45 +0,0 @@
-# Regression problem
-# Originally written by Keegan Go for lsqpy
-# Translated into LinearLeastSquares.jl by Karanveer Mohan and David Zeng
-
-"""
-Fit a large polynomial to some given data
-Play with regularization to understand
-Question: Why does increasing regularization not help very much
-near the end of the polynomial (farther from zero)?
-"""
-
-using LinearLeastSquares
-import PyPlot.plt
-
-# Set the random seed to get consistent data
-srand(1);
-
-# Number of examples to use
-n = 50
-
-# Generate data
-x_data = rand(n, 1) * 18;
-y_data = sin(x_data * 2) + cos(x_data)- 4 * cos(x_data) / 4 + 3 * sin(x_data / 3) + 0.2 * rand(n, 1);
-
-num_powers = 20;
-
-t_vals = 0 : 0.1 : 18;
-t = reshape(t_vals, length(t_vals), 1);
-T = hcat([t .^ i for i in 1 : num_powers]...);
-
-# We will regress using different powers of x
-X = hcat([x_data .^ i for i in 1 : num_powers]...);
-
-# Solve the problem
-mu = 0;
-a = Variable(num_powers);
-optval = minimize!(sum_squares(X * a - y_data) + mu * sum_squares(a));
-
-# Plot our regressed function
-plt.plot(x_data, y_data, "bo");
-plt.plot(t, T * a.value, "r");
-plt.xlabel("x");
-plt.ylabel("y");
-plt.ylim([-5, 5]);
-plt.show();
diff --git a/examples/regression/data.jl b/examples/regression/data.jl
new file mode 100644
index 0000000..b9b73de
--- /dev/null
+++ b/examples/regression/data.jl
@@ -0,0 +1,22 @@
+using Gadfly
+
+# Set the random seed to get consistent data
+srand(1)
+
+# Number of examples to use
+n = 100
+
+# Specify the true value of the variable
+true_coeffs = [2; -2; 0.5];
+
+# Generate data
+x_data = rand(n, 1) * 5;
+x_data_expanded = hcat([x_data .^ i for i in 1 : 3]...)
+y_data = x_data_expanded * true_coeffs + 0.5 * rand(n, 1)
+
+
+p = plot(
+  x=x_data, y=y_data, Geom.point,
+  Theme(panel_fill=color("white"))
+)
+draw(PNG("data.png", 16cm, 12cm), p)
diff --git a/examples/regression/data.png b/examples/regression/data.png
new file mode 100644
index 0000000..7ae7d09
Binary files /dev/null and b/examples/regression/data.png differ
diff --git a/examples/regression/linear_regression.jl b/examples/regression/linear_regression.jl
new file mode 100644
index 0000000..7a26a5e
--- /dev/null
+++ b/examples/regression/linear_regression.jl
@@ -0,0 +1,35 @@
+# Simple Linear Regression
+# Originally written by Keegan Go for lsqpy
+# Translated into LinearLeastSquares.jl by Karanveer Mohan and David Zeng
+
+using LinearLeastSquares
+using Gadfly
+# Set the random seed to get consistent data
+srand(1)
+
+# Number of examples to use
+n = 100
+
+# Specify the true value of the variable
+true_coeffs = [2; -2; 0.5]
+
+# Generate data
+x_data = rand(n) * 5
+x_data_expanded = hcat([x_data .^ i for i in 1 : 3]...)
+y_data = x_data_expanded * true_coeffs + 0.5 * rand(n)
+
+slope = Variable()
+offset = Variable()
+line = offset + x_data * slope
+residuals = line - y_data
+error = sum_squares(residuals)
+optval = minimize!(error)
+
+# plot the data and the line
+t = [0; 5; 0.1]
+p = plot(
+  layer(x=x_data, y=y_data, Geom.point),
+  layer(x=t, y=evaluate(slope) * t + evaluate(offset), Geom.line),
+  Theme(panel_fill=color("white"))
+)
+# draw(PNG("linear_regression.png", 16cm, 12cm), p)
diff --git a/examples/regression/linear_regression.png b/examples/regression/linear_regression.png
new file mode 100644
index 0000000..f20a214
Binary files /dev/null and b/examples/regression/linear_regression.png differ
diff --git a/examples/simple_lin_and_quad_reg/quadratic_regression.jl b/examples/regression/quadratic_regression.jl
similarity index 54%
rename from examples/simple_lin_and_quad_reg/quadratic_regression.jl
rename to examples/regression/quadratic_regression.jl
index a5ea0e5..0dbe8f1 100644
--- a/examples/simple_lin_and_quad_reg/quadratic_regression.jl
+++ b/examples/regression/quadratic_regression.jl
@@ -3,6 +3,7 @@
 # Translated into LinearLeastSquares.jl by Karanveer Mohan and David Zeng
 
 using LinearLeastSquares
+using Gadfly
 # Set the random seed to get consistent data
 srand(1)
 
@@ -10,28 +11,29 @@ srand(1)
 n = 100
 
 # Specify the true value of the variable
-true_coeffs = [2; -2; 0.5];
+true_coeffs = [2; -2; 0.5]
 
 # Generate data
 x_data = rand(n, 1) * 5
 x_data_expanded = hcat([x_data .^ i for i in 1 : 3]...)
 y_data = x_data_expanded * true_coeffs + 0.5 * rand(n, 1)
 
-quadratic_coeff = Variable();
-slope = Variable();
-offset = Variable();
-quadratic = offset + x_data * slope + quadratic_coeff * x_data .^ 2;
-residuals = quadratic - y_data;
-error = sum_squares(residuals);
-optval = minimize!(error);
+quadratic_coeff = Variable()
+slope = Variable()
+offset = Variable()
+quadratic = offset + x_data * slope + quadratic_coeff * x_data .^ 2
+residuals = quadratic - y_data
+error = sum_squares(residuals)
+optval = minimize!(error)
 
 # Create some evenly spaced points for plotting, again replicate powers
-t = reshape([0 : 0.1 : 5], length([0 : 0.1 : 5]), 1);
-t_squared = t .^ 2;
+t = reshape([0 : 0.1 : 5], length([0 : 0.1 : 5]), 1)
+t_squared = t .^ 2
 
 # Plot our regressed function
-plt.plot(x_data, y_data, "ro")
-plt.plot(t, evaluate(offset) + t * evaluate(slope) + t_squared * evaluate(quadratic_coeff), "b")
-plt.xlabel("x")
-plt.ylabel("y")
-plt.show()
+p = plot(
+  layer(x=x_data, y=y_data, Geom.point),
+  layer(x=t, y=evaluate(offset) + t * evaluate(slope) + t_squared * evaluate(quadratic_coeff), Geom.line),
+  Theme(panel_fill=color("white"))
+)
+# draw(PNG("quadratic_regression.png", 16cm, 12cm), p)
diff --git a/examples/regression/quadratic_regression.png b/examples/regression/quadratic_regression.png
new file mode 100644
index 0000000..5f8a561
Binary files /dev/null and b/examples/regression/quadratic_regression.png differ
diff --git a/examples/simple_classifier.jl b/examples/simple_classifier.jl
deleted file mode 100644
index 501b9cd..0000000
--- a/examples/simple_classifier.jl
+++ /dev/null
@@ -1,34 +0,0 @@
-# Simple Classifier
-# Originally written by Keegan Go for lsqpy
-# Translated into LinearLeastSquares.jl by Karanveer Mohan and David Zeng
-
-using LinearLeastSquares
-import PyPlot.plt
-
-# Set the random seed to get consistent data
-srand(1);
-
-# Number of examples to use
-n = 200;
-
-# Specify the true value of the variable
-true_vect = [-1; 1];
-
-# Create data and labels
-X = rand(n, 2) * 2;
-y = sign(X * true_vect + rand(n, 1));
-
-a = Variable(2);
-optval = minimize!(sum_squares(X * a - y));
-println(a.value);
-
-# Plot the line we found
-plt.plot(X[y[:, 1] .>= 0, 1], X[y[:, 1] .>= 0, 2], "ro");
-plt.plot(X[y[:, 1] .< 0, 1], X[y[:, 1] .< 0, 2], "bo");
-
-t = [-5 : 0.1 : 5];
-plt.plot(t, -a.value[1, 1] * t / a.value[2, 1]);
-plt.xlabel("x");
-plt.ylabel("y");
-plt.xlim([0, 2]);
-plt.ylim([0, 2]);
diff --git a/examples/simple_lin_and_quad_reg/data.jl b/examples/simple_lin_and_quad_reg/data.jl
deleted file mode 100644
index 174f83c..0000000
--- a/examples/simple_lin_and_quad_reg/data.jl
+++ /dev/null
@@ -1,18 +0,0 @@
-import PyPlot.plt
-
-# Set the random seed to get consistent data
-srand(1)
-
-# Number of examples to use
-n = 100
-
-# Specify the true value of the variable
-true_coeffs = [2; -2; 0.5];
-
-# Generate data
-x_data = rand(n, 1) * 5;
-x_data_expanded = hcat([x_data .^ i for i in 1 : 3]...);
-y_data = x_data_expanded * true_coeffs + 0.5 * rand(n, 1);
-
-plt.plot(x_data, y_data, "ro");
-plt.show()
diff --git a/examples/simple_lin_and_quad_reg/data.png b/examples/simple_lin_and_quad_reg/data.png
deleted file mode 100644
index 1f124f0..0000000
Binary files a/examples/simple_lin_and_quad_reg/data.png and /dev/null differ
diff --git a/examples/simple_lin_and_quad_reg/linear.png b/examples/simple_lin_and_quad_reg/linear.png
deleted file mode 100644
index d81f398..0000000
Binary files a/examples/simple_lin_and_quad_reg/linear.png and /dev/null differ
diff --git a/examples/simple_lin_and_quad_reg/linear_regression.jl b/examples/simple_lin_and_quad_reg/linear_regression.jl
deleted file mode 100644
index 2507014..0000000
--- a/examples/simple_lin_and_quad_reg/linear_regression.jl
+++ /dev/null
@@ -1,34 +0,0 @@
-# Simple Linear Regression
-# Originally written by Keegan Go for lsqpy
-# Translated into LinearLeastSquares.jl by Karanveer Mohan and David Zeng
-
-using LinearLeastSquares
-import PyPlot.plt
-# Set the random seed to get consistent data
-srand(1)
-
-# Number of examples to use
-n = 100
-
-# Specify the true value of the variable
-true_coeffs = [2; -2; 0.5];
-
-# Generate data
-x_data = rand(n, 1) * 5;
-x_data_expanded = hcat([x_data .^ i for i in 1 : 3]...);
-y_data = x_data_expanded * true_coeffs + 0.5 * rand(n, 1);
-
-slope = Variable();
-offset = Variable();
-line = offset + x_data * slope;
-residuals = line - y_data;
-error = sum_squares(residuals);
-optval = minimize!(error);
-
-# plot the data and the line
-t = [0; 5; 0.1];
-plt.plot(x_data, y_data, "ro");
-plt.plot(t, evaluate(slope) .* t .+ evaluate(offset));
-plt.xlabel("x");
-plt.ylabel("y");
-plt.show();
diff --git a/examples/simple_lin_and_quad_reg/quadratic.png b/examples/simple_lin_and_quad_reg/quadratic.png
deleted file mode 100644
index 01eb799..0000000
Binary files a/examples/simple_lin_and_quad_reg/quadratic.png and /dev/null differ
diff --git a/examples/temperature/CRUTEM4v-gl.dat b/examples/temperature/CRUTEM4v-gl.dat
deleted file mode 100644
index 6c2b99e..0000000
--- a/examples/temperature/CRUTEM4v-gl.dat
+++ /dev/null
@@ -1,328 +0,0 @@
-1851 0.757 0.333 -0.511 0.464 0.056 0.079 0.342 0.277 0.196 0.643 -0.109 -0.181 0.196
-1851 3 3 3 3 3 3 3 3 3 3 3 3
-1852 0.166 0.186 -0.042 -0.813 0.409 0.459 0.517 0.095 0.166 -0.201 0.126 1.686 0.230
-1852 3 3 3 3 3 3 3 3 3 3 3 3
-1853 0.675 -0.596 -1.249 -0.460 -0.194 -0.174 0.362 0.282 -0.317 0.062 -0.500 -1.150 -0.272
-1853 3 3 3 3 3 3 3 3 3 3 3 3
-1854 -0.427 0.207 0.175 -0.330 0.383 0.082 0.552 0.334 0.131 0.617 -0.487 0.310 0.129
-1854 3 3 3 3 3 3 3 3 3 3 3 3
-1855 0.020 -1.295 -0.523 0.379 0.180 0.142 0.221 0.083 -0.301 0.287 -0.431 -1.540 -0.231
-1855 3 3 3 3 3 3 3 3 3 3 3 3
-1856 -0.277 -0.713 -1.209 -0.061 -0.574 0.183 -0.395 -0.300 -0.575 -0.572 -1.397 -0.153 -0.504
-1856 4 4 3 3 3 3 3 3 3 3 3 3
-1857 -1.081 0.203 -0.685 -1.363 -0.964 -0.320 -0.036 0.101 -0.329 -0.398 -0.717 0.560 -0.419
-1857 3 4 3 4 4 4 4 4 4 4 4 4
-1858 -0.003 -1.310 -0.723 -0.395 -0.377 0.111 -0.168 -0.348 -0.351 -0.080 -1.719 -0.371 -0.478
-1858 4 4 4 4 4 4 4 4 4 4 4 4
-1859 -0.156 0.148 -0.034 -0.176 -0.021 -0.187 0.018 0.008 -0.535 -0.180 -0.316 -1.036 -0.206
-1859 4 4 4 4 4 4 4 4 4 4 4 4
-1860 0.106 -1.067 -1.486 -0.533 -0.296 0.001 -0.280 -0.068 -0.282 -0.348 -1.218 -1.536 -0.584
-1860 4 4 4 4 4 4 4 4 4 4 4 4
-1861 -1.899 0.202 0.048 -0.774 -0.980 0.113 -0.141 0.029 -0.323 -0.277 -0.691 -0.438 -0.428
-1861 4 4 4 4 4 4 4 4 4 4 4 4
-1862 -1.221 -1.272 -0.444 -0.318 -0.180 -0.395 -0.457 -0.702 -0.367 -0.517 -0.972 -1.115 -0.663
-1862 4 4 4 3 3 3 4 4 4 4 4 4
-1863 1.068 0.418 -0.369 -0.308 -0.230 -0.381 -0.471 -0.255 -0.470 -0.578 -0.367 -0.363 -0.192
-1863 3 4 4 4 4 4 4 3 3 3 3 3
-1864 -1.350 -0.583 -0.494 -0.738 -0.621 0.097 -0.077 -0.477 -0.304 -1.297 -1.263 -1.225 -0.694
-1864 4 4 4 4 4 4 4 4 4 4 4 4
-1865 -0.131 -1.306 -1.385 -0.179 -0.168 -0.211 0.106 -0.351 0.236 -0.595 0.023 -0.542 -0.375
-1865 4 4 4 4 4 4 4 4 4 4 4 4
-1866 0.547 -0.509 -0.897 -0.226 -0.729 0.034 -0.037 -0.506 -0.042 -0.818 -0.628 -0.374 -0.349
-1866 4 4 4 4 4 4 4 4 4 4 4 4
-1867 -0.711 0.560 -1.827 -0.492 -1.138 -0.270 -0.341 -0.184 -0.104 -0.109 -0.631 -0.962 -0.517
-1867 4 4 4 4 4 4 4 4 4 4 4 4
-1868 -1.348 -0.833 -0.208 -0.664 0.159 0.089 0.423 0.098 -0.354 -0.153 -0.980 -0.203 -0.331
-1868 4 4 4 5 5 4 5 5 5 4 5 5
-1869 -0.166 0.827 -1.125 -0.352 -0.444 -0.345 -0.167 0.192 -0.071 -0.972 -0.613 -0.415 -0.304
-1869 5 5 5 5 5 5 4 4 5 5 4 4
-1870 0.103 -0.959 -1.112 -0.446 -0.018 0.185 0.123 -0.704 -0.318 -0.482 -0.149 -1.758 -0.461
-1870 5 5 5 5 5 5 5 5 5 5 5 5
-1871 -0.997 -1.155 -0.007 -0.335 -0.556 -0.396 0.008 0.110 -0.593 -0.604 -1.268 -1.203 -0.583
-1871 5 5 5 5 5 5 5 5 5 5 5 5
-1872 -0.124 -0.498 -0.624 -0.124 -0.044 -0.089 -0.132 -0.422 -0.194 -0.177 -0.504 -0.789 -0.310
-1872 5 5 6 5 6 6 5 5 5 5 6 5
-1873 -0.031 -0.629 -0.530 -0.988 -0.587 0.203 -0.155 0.004 -0.398 -0.422 -0.918 -0.020 -0.373
-1873 6 6 6 6 6 6 6 6 6 6 6 6
-1874 0.256 -0.469 -1.071 -0.670 -0.840 -0.404 -0.248 -0.388 -0.201 -0.250 -0.806 -0.388 -0.457
-1874 6 6 6 6 6 6 6 6 6 6 6 6
-1875 -1.212 -1.581 -1.624 -1.047 -0.254 -0.195 -0.324 -0.298 -0.640 -0.814 -1.263 -0.888 -0.845
-1875 6 7 7 6 7 7 7 7 7 7 7 7
-1876 -0.270 -0.228 -0.458 -0.374 -0.745 -0.082 0.090 -0.172 -0.216 -0.446 -1.265 -1.450 -0.468
-1876 7 7 7 7 7 7 7 7 7 7 7 7
-1877 -0.317 -0.046 -0.637 -0.739 -0.885 -0.069 0.146 0.185 -0.384 -0.561 0.058 0.098 -0.263
-1877 7 7 7 7 7 7 7 7 7 7 7 7
-1878 -0.342 0.404 0.598 0.369 -0.435 -0.114 -0.095 -0.026 0.240 0.107 -0.155 -0.965 -0.034
-1878 7 7 7 7 7 7 7 7 7 7 7 7
-1879 -0.483 -0.482 -0.455 -0.625 -0.527 -0.587 -0.377 -0.368 -0.346 -0.124 -0.876 -1.181 -0.536
-1879 8 8 8 8 8 8 8 8 8 8 8 8
-1880 0.086 -0.266 -0.499 -0.343 -0.110 -0.128 -0.201 0.121 -0.181 -0.975 -1.041 -0.616 -0.346
-1880 8 8 8 8 8 8 8 8 8 8 8 8
-1881 -1.244 -0.620 -0.721 -0.477 -0.054 -0.602 -0.294 -0.163 -0.148 -0.666 -0.689 0.092 -0.465
-1881 8 8 8 9 9 9 9 8 8 9 8 8
-1882 0.598 0.080 0.099 -0.693 -0.720 -0.664 -0.538 -0.118 -0.122 -0.391 -0.758 -0.894 -0.343
-1882 9 9 9 9 9 9 9 9 9 9 9 9
-1883 -1.072 -0.913 -1.105 -0.710 -0.566 0.011 -0.279 -0.449 -0.548 -0.561 -0.573 -0.362 -0.594
-1883 9 9 9 9 9 9 9 9 9 9 9 9
-1884 -0.601 -0.415 -0.808 -1.056 -0.993 -0.684 -0.717 -0.269 -0.421 -0.515 -0.996 -0.637 -0.676
-1884 9 9 9 9 9 9 9 9 9 9 9 9
-1885 -1.315 -0.794 -1.034 -0.883 -0.882 -0.703 -0.367 -0.633 -0.373 -0.526 -0.462 -0.012 -0.665
-1885 9 9 9 9 9 9 9 9 9 9 10 10
-1886 -0.785 -1.042 -0.964 -0.311 -0.413 -0.647 -0.337 -0.399 -0.225 -0.567 -0.494 -0.302 -0.541
-1886 10 10 10 10 10 10 10 10 10 10 10 10
-1887 -0.712 -0.779 -0.577 -0.658 -0.434 -0.425 -0.299 -0.464 -0.315 -0.879 -0.669 -0.610 -0.569
-1887 10 10 10 10 10 10 10 10 10 10 10 10
-1888 -1.170 -1.003 -1.434 -0.344 -0.747 -0.469 -0.417 -0.452 -0.241 -0.416 -0.498 -0.318 -0.626
-1888 10 11 10 11 11 10 10 10 10 10 11 11
-1889 -0.446 -0.458 -0.304 -0.012 0.108 -0.171 -0.368 -0.400 -0.566 -0.353 -0.693 0.024 -0.303
-1889 11 11 11 11 11 11 11 11 11 11 11 11
-1890 -0.152 -0.372 -0.509 -0.270 -0.617 -0.299 -0.433 -0.440 -0.263 -0.531 -0.801 -0.705 -0.449
-1890 11 11 11 11 11 11 11 11 11 11 11 11
-1891 -0.935 -0.834 -0.691 -0.721 -0.474 -0.502 -0.519 -0.529 -0.232 -0.513 -0.922 0.160 -0.559
-1891 12 12 12 12 12 12 12 12 12 12 12 11
-1892 -0.726 -0.082 -0.884 -0.865 -0.805 -0.436 -0.494 -0.361 -0.280 -0.478 -0.784 -1.325 -0.627
-1892 12 12 12 12 12 12 12 12 12 12 12 12
-1893 -2.046 -1.506 -0.603 -0.677 -0.759 -0.487 -0.205 -0.390 -0.496 -0.265 -0.535 -0.387 -0.696
-1893 12 12 12 12 12 12 12 12 12 12 12 12
-1894 -0.552 -0.457 -0.212 -0.419 -0.538 -0.591 -0.425 -0.391 -0.741 -0.439 -0.443 -0.411 -0.468
-1894 12 12 12 12 12 12 12 12 12 12 12 12
-1895 -1.076 -1.525 -0.761 -0.458 -0.488 -0.377 -0.652 -0.356 -0.183 -0.345 -0.459 -0.277 -0.580
-1895 12 12 12 12 12 12 12 13 12 13 13 13
-1896 -0.348 -0.353 -0.842 -0.797 -0.371 -0.256 -0.244 -0.342 -0.313 -0.055 -0.950 -0.142 -0.418
-1896 13 13 13 13 13 13 13 13 13 13 13 13
-1897 -0.599 -0.375 -0.802 -0.089 -0.059 -0.172 -0.131 -0.273 -0.063 -0.294 -0.502 -0.580 -0.328
-1897 13 13 13 13 13 13 13 13 13 13 13 13
-1898 0.324 -0.394 -1.290 -0.659 -0.613 -0.213 -0.330 -0.134 -0.134 -0.570 -0.416 -0.135 -0.380
-1898 14 14 14 14 14 14 14 14 14 14 14 14
-1899 -0.124 -0.730 -0.809 -0.245 -0.466 -0.554 -0.266 -0.224 -0.005 -0.199 0.318 -0.610 -0.326
-1899 14 14 14 14 14 14 14 14 14 14 14 14
-1900 -0.397 -0.426 -0.470 -0.409 -0.279 -0.030 -0.284 -0.157 -0.236 0.288 -0.193 0.139 -0.204
-1900 14 14 14 14 14 14 14 14 14 14 14 14
-1901 -0.179 -0.545 -0.178 -0.070 -0.144 0.057 -0.020 -0.090 -0.152 -0.178 -0.329 -0.457 -0.190
-1901 14 14 14 14 14 14 14 14 14 14 14 14
-1902 0.045 -0.102 -0.311 -0.480 -0.455 -0.371 -0.445 -0.497 -0.465 -0.542 -0.604 -0.842 -0.422
-1902 15 15 15 15 15 15 14 15 15 15 15 15
-1903 -0.115 0.110 -0.120 -0.523 -0.556 -0.583 -0.527 -0.488 -0.400 -0.556 -0.591 -0.732 -0.424
-1903 15 15 15 15 15 15 15 15 15 15 15 15
-1904 -0.783 -0.696 -0.883 -0.535 -0.523 -0.505 -0.538 -0.467 -0.572 -0.389 -0.283 -0.484 -0.555
-1904 15 15 15 15 15 15 15 15 15 15 15 15
-1905 -0.568 -1.229 -0.518 -0.696 -0.397 -0.231 -0.336 -0.345 -0.250 -0.544 -0.067 -0.029 -0.434
-1905 15 15 15 15 16 15 15 15 15 15 15 15
-1906 0.100 -0.471 -0.604 0.122 -0.023 -0.134 -0.194 -0.216 -0.333 -0.252 -0.621 -0.154 -0.232
-1906 15 16 16 16 16 15 15 16 15 15 16 16
-1907 -0.700 -0.785 -0.528 -0.917 -0.967 -0.706 -0.496 -0.602 -0.420 -0.147 -0.782 -0.546 -0.633
-1907 16 16 16 16 16 16 16 16 16 16 16 16
-1908 -0.207 -0.393 -0.867 -0.567 -0.457 -0.398 -0.418 -0.576 -0.314 -0.595 -0.726 -0.410 -0.494
-1908 16 16 16 16 16 16 16 16 16 16 16 16
-1909 -0.631 -0.661 -0.872 -0.838 -0.797 -0.408 -0.437 -0.078 -0.152 -0.139 -0.091 -0.686 -0.482
-1909 16 16 16 16 16 16 16 16 17 16 16 16
-1910 -0.094 -0.377 -0.244 -0.224 -0.347 -0.312 -0.222 -0.259 -0.224 -0.413 -0.786 -0.662 -0.347
-1910 17 17 17 17 17 17 17 17 17 17 17 17
-1911 -0.587 -0.914 -0.799 -0.715 -0.410 -0.434 -0.282 -0.264 -0.296 -0.388 -0.202 -0.075 -0.447
-1911 17 18 18 18 18 17 17 17 17 18 17 18
-1912 -0.485 -0.099 -0.588 -0.442 -0.331 -0.144 -0.603 -0.693 -0.686 -0.810 -0.572 -0.275 -0.477
-1912 18 18 18 18 18 18 18 18 18 18 18 18
-1913 -0.257 -0.723 -0.638 -0.244 -0.828 -0.669 -0.379 -0.211 -0.339 -0.372 0.187 0.271 -0.350
-1913 18 18 18 18 18 18 18 18 18 18 18 18
-1914 0.454 0.138 -0.149 -0.309 -0.127 -0.115 -0.195 -0.155 -0.211 -0.000 -0.336 -0.344 -0.112
-1914 18 18 18 18 18 18 17 18 18 18 18 18
-1915 -0.088 0.302 -0.660 0.063 -0.340 -0.242 -0.034 -0.215 -0.075 -0.250 -0.034 -0.106 -0.140
-1915 17 17 17 17 17 17 17 17 17 17 17 17
-1916 0.017 -0.115 -0.726 -0.357 -0.381 -0.525 -0.227 -0.259 -0.330 -0.381 -0.466 -0.930 -0.390
-1916 18 18 18 18 18 18 17 17 18 18 18 18
-1917 -0.629 -1.340 -1.214 -0.765 -1.183 -0.440 -0.107 -0.330 -0.173 -0.609 -0.333 -1.267 -0.699
-1917 18 18 18 18 18 18 18 18 18 18 18 18
-1918 -0.788 -0.565 -0.510 -0.614 -0.736 -0.415 -0.466 -0.371 -0.306 0.121 -0.302 -0.319 -0.439
-1918 18 18 18 18 18 18 18 18 18 18 18 17
-1919 -0.080 -0.317 -0.599 -0.115 -0.467 -0.145 -0.234 -0.200 -0.072 -0.294 -0.835 -0.559 -0.326
-1919 17 17 17 17 17 17 17 17 17 17 17 17
-1920 -0.130 -0.417 -0.112 -0.346 -0.210 -0.280 -0.157 -0.225 -0.106 -0.374 -0.546 -0.495 -0.283
-1920 17 17 17 17 17 18 17 17 17 17 17 17
-1921 0.357 -0.118 -0.005 -0.033 0.053 0.064 -0.024 -0.369 -0.199 -0.211 -0.609 -0.145 -0.103
-1921 18 18 18 18 18 18 18 18 18 19 19 19
-1922 -0.631 -0.513 -0.367 -0.117 -0.165 -0.163 -0.281 -0.274 -0.202 -0.286 -0.161 -0.191 -0.279
-1922 19 19 19 19 19 19 19 19 19 19 19 19
-1923 0.071 -0.602 -0.376 -0.709 -0.377 -0.298 -0.433 -0.467 -0.189 -0.060 0.137 0.200 -0.259
-1923 19 19 19 19 19 19 19 19 19 19 20 19
-1924 -0.488 -0.492 -0.497 -0.630 -0.331 -0.207 -0.212 -0.237 -0.091 -0.166 -0.146 -0.583 -0.340
-1924 20 20 20 20 20 19 20 20 20 20 20 20
-1925 -0.327 -0.055 -0.287 -0.125 -0.395 -0.314 -0.395 -0.120 -0.245 -0.541 -0.084 0.170 -0.226
-1925 20 20 20 20 20 20 20 20 20 20 20 20
-1926 0.388 0.475 0.017 -0.372 -0.509 -0.211 -0.250 -0.077 -0.042 -0.089 0.074 -0.447 -0.087
-1926 20 20 20 20 20 20 20 20 20 20 20 20
-1927 -0.267 -0.174 -0.392 -0.429 -0.501 -0.259 -0.079 -0.171 0.020 0.215 -0.067 -0.860 -0.247
-1927 20 20 20 20 20 20 20 20 20 20 20 20
-1928 0.099 -0.113 -0.501 -0.305 -0.341 -0.569 -0.107 -0.130 -0.042 -0.069 0.104 -0.007 -0.165
-1928 20 20 20 20 21 20 20 20 20 20 20 20
-1929 -0.700 -1.366 -0.481 -0.695 -0.443 -0.431 -0.417 -0.164 -0.277 0.023 0.009 -0.699 -0.470
-1929 21 21 21 21 21 21 21 21 21 21 21 21
-1930 -0.347 -0.122 -0.057 -0.151 -0.315 -0.053 -0.007 0.037 -0.162 -0.219 0.104 -0.115 -0.117
-1930 21 21 21 21 21 21 21 21 21 21 21 21
-1931 0.024 -0.414 -0.330 -0.402 -0.342 -0.001 0.093 0.046 -0.053 0.154 -0.116 0.121 -0.102
-1931 21 21 21 22 22 22 22 22 22 22 21 21
-1932 0.878 -0.422 -0.676 0.030 -0.221 -0.195 -0.084 -0.041 0.080 0.057 -0.272 -0.147 -0.084
-1932 22 22 22 22 22 22 22 22 22 22 22 22
-1933 -0.450 -0.639 -0.541 -0.297 -0.257 -0.230 -0.055 -0.080 -0.072 -0.034 -0.365 -0.874 -0.325
-1933 22 22 23 23 23 23 23 23 23 23 23 23
-1934 -0.005 0.201 -0.399 -0.350 0.217 -0.036 0.014 -0.014 -0.141 0.035 0.219 0.122 -0.011
-1934 23 23 23 23 23 23 23 23 23 23 23 23
-1935 -0.435 0.723 -0.071 -0.563 -0.483 -0.200 -0.064 -0.040 -0.144 0.124 -0.508 -0.305 -0.164
-1935 23 23 23 23 23 23 23 23 23 23 23 23
-1936 -0.311 -0.909 -0.366 -0.326 -0.118 -0.023 0.282 0.114 -0.014 0.031 0.063 0.255 -0.110
-1936 24 24 24 24 24 24 24 24 24 24 24 24
-1937 -0.253 0.105 -0.689 -0.314 -0.029 0.085 0.002 0.215 0.245 0.322 0.112 -0.301 -0.042
-1937 24 24 24 24 24 24 24 24 24 24 24 24
-1938 0.157 0.162 0.369 0.331 0.045 -0.034 0.076 0.113 0.274 0.448 0.336 -0.357 0.160
-1938 24 24 24 24 24 24 24 24 24 24 24 24
-1939 0.128 0.058 -0.464 -0.175 0.052 0.109 -0.022 0.069 -0.139 -0.282 -0.008 0.697 0.002
-1939 25 25 25 25 25 25 25 25 25 25 25 25
-1940 -0.796 -0.111 -0.151 0.006 -0.160 0.152 0.119 0.024 0.158 -0.029 -0.038 0.279 -0.046
-1940 25 25 25 25 25 25 25 25 25 25 25 25
-1941 -0.212 0.085 -0.377 -0.036 -0.049 0.107 0.174 0.021 -0.190 0.192 -0.204 -0.164 -0.054
-1941 26 26 26 26 26 26 26 26 26 26 26 26
-1942 0.076 -0.428 -0.294 -0.111 -0.110 0.082 -0.099 -0.008 0.104 0.103 0.071 0.033 -0.049
-1942 27 27 27 27 27 27 26 27 27 27 27 27
-1943 -0.375 0.286 -0.247 0.176 0.056 -0.243 -0.003 -0.148 0.040 0.386 -0.010 0.283 0.017
-1943 27 27 27 27 27 27 27 27 27 27 27 27
-1944 0.719 0.299 0.035 -0.146 -0.042 -0.025 0.049 0.139 0.397 0.325 -0.055 -0.358 0.111
-1944 27 27 27 27 27 27 27 27 27 27 27 27
-1945 -0.256 -0.426 -0.186 0.189 -0.275 -0.118 -0.187 0.398 0.136 0.125 -0.130 -0.674 -0.117
-1945 27 27 27 27 27 27 27 27 27 27 27 27
-1946 0.121 0.134 -0.291 0.184 -0.098 -0.246 0.007 -0.026 0.053 -0.162 0.006 -0.553 -0.073
-1946 28 28 27 27 27 27 27 27 27 28 28 27
-1947 -0.293 -0.332 -0.003 0.147 -0.039 0.050 -0.056 0.076 0.060 0.413 0.072 -0.236 -0.011
-1947 28 28 28 28 28 28 28 28 28 28 28 28
-1948 0.416 -0.090 -0.399 -0.008 0.156 0.124 -0.104 -0.051 0.019 0.139 0.047 -0.223 0.002
-1948 28 28 29 29 29 29 28 28 28 29 29 29
-1949 0.366 -0.338 -0.392 -0.098 -0.110 -0.240 -0.173 -0.045 -0.084 0.125 -0.011 -0.349 -0.112
-1949 29 29 29 29 29 29 29 29 29 29 29 29
-1950 -0.626 -0.366 -0.244 -0.294 -0.151 -0.093 -0.230 -0.284 -0.145 -0.184 -0.670 -0.308 -0.300
-1950 30 30 30 31 31 30 30 31 31 31 31 30
-1951 -0.517 -0.795 -0.427 -0.078 -0.031 -0.102 -0.046 0.108 0.187 0.133 -0.033 0.393 -0.101
-1951 34 35 35 35 35 34 35 35 35 35 35 35
-1952 0.278 0.063 -0.408 -0.036 -0.054 -0.020 0.104 0.078 0.110 -0.147 -0.540 -0.208 -0.065
-1952 35 35 35 35 35 35 35 35 35 35 35 35
-1953 0.153 0.170 0.201 0.201 0.038 0.162 -0.018 0.130 0.099 0.181 -0.148 0.241 0.117
-1953 36 35 36 36 36 36 36 36 36 36 36 36
-1954 -0.494 -0.227 -0.361 -0.204 -0.297 -0.101 -0.178 -0.062 -0.004 0.071 0.231 -0.186 -0.151
-1954 36 36 36 36 36 36 36 36 36 36 36 36
-1955 0.393 -0.130 -0.773 -0.346 -0.196 -0.107 -0.200 0.033 -0.035 0.006 -0.373 -0.480 -0.184
-1955 36 37 37 37 37 37 37 37 36 37 36 36
-1956 -0.322 -0.686 -0.560 -0.518 -0.413 -0.249 -0.283 -0.371 -0.401 -0.305 -0.466 -0.359 -0.411
-1956 37 37 37 37 37 37 37 37 37 37 37 37
-1957 -0.265 -0.217 -0.449 -0.167 -0.114 0.055 -0.175 0.025 0.049 -0.066 0.130 0.345 -0.071
-1957 37 37 37 37 37 37 37 37 37 37 37 37
-1958 0.480 0.227 -0.088 0.008 0.040 -0.083 0.009 -0.040 -0.127 0.004 0.075 -0.016 0.041
-1958 37 37 37 37 37 37 37 37 37 37 37 37
-1959 0.120 0.057 0.193 0.060 -0.120 0.117 0.047 0.023 0.046 -0.164 -0.176 -0.017 0.016
-1959 38 37 38 37 38 37 38 38 38 38 37 37
-1960 0.029 0.333 -0.693 -0.292 -0.322 0.052 -0.092 -0.021 0.037 0.011 -0.213 0.299 -0.073
-1960 37 38 37 37 37 38 38 37 38 38 38 37
-1961 0.090 0.278 0.106 0.092 0.077 0.083 -0.034 -0.009 0.046 -0.050 0.006 -0.178 0.042
-1961 38 38 38 38 38 39 38 38 38 38 38 38
-1962 0.186 0.263 -0.018 0.046 -0.083 -0.145 -0.076 -0.071 -0.021 0.111 0.071 0.042 0.025
-1962 39 39 39 38 39 39 39 39 39 39 39 39
-1963 -0.086 0.388 -0.309 -0.228 -0.170 -0.135 0.059 0.042 0.132 0.384 0.266 -0.133 0.018
-1963 39 39 39 39 39 39 39 39 39 39 39 39
-1964 -0.036 -0.288 -0.515 -0.369 -0.178 -0.235 -0.194 -0.275 -0.279 -0.394 -0.282 -0.426 -0.289
-1964 39 39 39 39 39 39 39 39 39 39 39 39
-1965 -0.066 -0.443 -0.282 -0.455 -0.225 -0.106 -0.283 -0.218 -0.202 -0.062 -0.257 -0.151 -0.229
-1965 39 39 39 39 39 39 39 39 39 39 39 39
-1966 -0.230 -0.169 -0.134 -0.297 -0.193 0.032 -0.015 -0.028 -0.035 -0.180 -0.163 -0.391 -0.150
-1966 39 39 39 39 39 39 39 39 39 39 39 39
-1967 -0.276 -0.434 -0.097 -0.079 0.099 -0.242 -0.121 -0.091 -0.132 0.333 0.014 -0.124 -0.096
-1967 39 39 39 39 39 39 39 39 39 39 39 39
-1968 -0.359 -0.249 0.351 -0.203 -0.316 -0.294 -0.217 -0.203 -0.206 -0.134 -0.257 -0.441 -0.211
-1968 39 39 39 39 39 39 39 39 39 39 39 39
-1969 -0.648 -0.734 -0.331 -0.028 -0.035 -0.108 0.028 -0.033 -0.092 -0.070 0.171 0.226 -0.138
-1969 39 39 39 39 39 39 39 39 39 39 39 39
-1970 -0.068 0.279 -0.266 0.001 -0.113 0.076 -0.014 -0.078 0.003 -0.182 -0.118 -0.318 -0.067
-1970 39 39 39 39 39 39 39 39 39 39 39 39
-1971 -0.063 -0.421 -0.473 -0.365 -0.289 -0.301 -0.167 -0.108 -0.001 -0.064 0.006 -0.079 -0.194
-1971 38 38 38 38 38 38 38 38 39 38 38 38
-1972 -0.779 -0.630 -0.280 -0.168 -0.185 -0.039 -0.082 -0.083 -0.232 -0.108 -0.280 0.057 -0.234
-1972 38 38 38 38 38 38 38 38 38 38 38 38
-1973 0.150 0.505 0.362 0.245 0.140 0.237 0.069 0.021 -0.049 -0.011 -0.138 -0.062 0.122
-1973 38 38 38 38 38 38 38 38 38 38 38 38
-1974 -0.460 -0.605 -0.194 -0.155 -0.234 -0.188 -0.079 -0.120 -0.246 -0.347 -0.319 -0.326 -0.273
-1974 38 38 38 38 38 38 38 38 38 38 38 38
-1975 0.133 -0.034 0.028 0.030 0.015 -0.005 -0.018 -0.139 0.049 -0.173 -0.271 -0.261 -0.054
-1975 38 38 38 38 38 38 38 38 38 38 38 38
-1976 -0.060 -0.339 -0.718 -0.168 -0.307 -0.274 -0.297 -0.277 -0.222 -0.691 -0.447 -0.296 -0.341
-1976 38 38 38 38 38 38 38 38 38 38 38 38
-1977 -0.441 0.101 0.259 0.261 0.105 0.201 0.063 -0.009 0.018 -0.056 0.381 0.023 0.075
-1977 38 38 38 38 38 38 38 38 38 38 38 38
-1978 0.045 -0.046 0.136 0.003 -0.076 -0.165 -0.075 -0.258 -0.069 -0.060 0.149 -0.172 -0.049
-1978 38 38 38 38 38 38 38 38 38 38 38 38
-1979 -0.157 -0.301 0.019 -0.261 -0.076 0.097 0.028 0.054 0.088 0.187 0.124 0.630 0.036
-1979 38 38 38 38 38 38 38 38 38 38 38 38
-1980 0.101 0.252 -0.113 0.168 0.243 0.139 0.063 0.082 0.104 0.122 0.266 0.139 0.131
-1980 38 38 38 38 38 38 38 38 38 38 38 38
-1981 0.756 0.611 0.506 0.415 0.118 0.150 0.099 0.252 0.113 0.034 0.158 0.508 0.310
-1981 37 37 37 37 37 37 37 37 37 37 37 37
-1982 -0.200 0.102 -0.242 0.072 0.032 -0.114 0.041 0.027 0.106 -0.116 -0.093 0.461 0.006
-1982 36 36 36 36 36 36 36 37 37 37 37 37
-1983 0.744 0.490 0.368 0.138 0.142 0.107 0.246 0.333 0.379 0.216 0.560 0.139 0.322
-1983 36 37 37 37 37 37 37 37 37 37 37 37
-1984 0.179 -0.076 0.020 -0.133 0.218 0.035 0.005 0.044 -0.125 0.084 -0.203 -0.505 -0.038
-1984 37 36 37 37 37 37 37 37 36 36 37 37
-1985 -0.003 -0.309 -0.087 0.017 0.073 -0.051 -0.115 0.120 -0.020 0.048 -0.044 0.035 -0.028
-1985 37 37 37 37 37 37 37 37 37 37 37 37
-1986 0.448 0.192 0.237 0.293 0.203 0.161 -0.030 0.032 -0.053 0.011 -0.024 0.036 0.126
-1986 37 37 37 37 37 36 37 36 37 37 37 37
-1987 0.170 0.662 -0.118 0.183 0.171 0.212 0.356 0.184 0.298 0.227 0.098 0.516 0.247
-1987 37 37 37 37 37 37 37 37 37 37 37 37
-1988 0.621 0.250 0.402 0.324 0.360 0.418 0.342 0.326 0.381 0.373 0.122 0.507 0.369
-1988 37 37 37 37 37 36 36 37 36 37 37 36
-1989 0.180 0.455 0.424 0.249 0.191 0.167 0.222 0.220 0.217 0.304 0.076 0.314 0.252
-1989 36 36 36 36 36 36 36 36 36 36 36 36
-1990 0.393 0.536 1.160 0.603 0.400 0.418 0.283 0.351 0.196 0.477 0.571 0.413 0.483
-1990 36 36 36 36 36 36 36 36 36 36 36 36
-1991 0.510 0.523 0.247 0.581 0.328 0.566 0.438 0.364 0.381 0.362 0.260 0.117 0.390
-1991 35 34 34 34 33 34 33 34 34 34 33 33
-1992 0.723 0.568 0.431 0.080 0.173 0.063 -0.224 -0.058 -0.204 -0.192 -0.218 0.192 0.111
-1992 34 33 33 33 34 34 34 34 34 33 34 34
-1993 0.495 0.458 0.366 0.214 0.261 0.205 0.178 0.129 -0.109 0.111 -0.352 0.251 0.184
-1993 34 34 34 34 33 33 34 34 34 33 33 34
-1994 0.351 -0.176 0.338 0.424 0.387 0.440 0.280 0.278 0.350 0.508 0.500 0.420 0.342
-1994 34 34 34 34 34 34 34 34 34 35 35 34
-1995 0.879 1.223 0.511 0.513 0.362 0.522 0.423 0.605 0.477 0.678 0.572 0.221 0.582
-1995 34 34 35 35 34 35 34 35 34 35 34 34
-1996 0.202 0.455 0.138 0.008 0.307 0.284 0.303 0.242 0.090 0.184 0.245 0.411 0.239
-1996 34 34 34 34 34 34 34 34 34 34 34 34
-1997 0.428 0.657 0.554 0.422 0.359 0.528 0.405 0.501 0.576 0.656 0.535 0.626 0.521
-1997 34 34 34 34 34 33 34 34 33 34 34 33
-1998 0.722 1.318 0.768 0.927 0.810 0.858 0.954 0.894 0.687 0.682 0.275 0.854 0.812
-1998 34 34 34 34 34 34 34 35 34 34 34 34
-1999 0.767 1.163 0.218 0.492 0.404 0.476 0.514 0.445 0.573 0.555 0.390 0.819 0.568
-1999 34 34 34 34 34 34 34 34 34 34 34 34
-2000 0.347 0.984 0.778 0.946 0.516 0.448 0.405 0.560 0.415 0.292 0.073 0.204 0.497
-2000 34 34 34 34 34 34 34 34 34 34 33 33
-2001 0.725 0.520 0.789 0.673 0.706 0.576 0.652 0.804 0.556 0.612 1.019 0.542 0.681
-2001 31 31 32 32 31 31 31 32 32 32 31 31
-2002 1.198 1.358 1.144 0.692 0.600 0.664 0.784 0.624 0.697 0.498 0.734 0.278 0.773
-2002 32 32 33 32 32 32 32 32 33 33 33 33
-2003 0.988 0.573 0.578 0.613 0.845 0.713 0.640 0.799 0.753 0.915 0.668 1.113 0.766
-2003 33 32 32 32 32 33 32 32 32 33 32 33
-2004 0.805 1.060 0.911 0.706 0.371 0.543 0.422 0.481 0.573 0.775 1.065 0.475 0.682
-2004 33 33 33 32 32 32 32 32 32 32 32 32
-2005 0.957 0.562 0.853 1.009 0.762 0.867 0.829 0.716 0.902 1.077 1.186 0.782 0.875
-2005 32 32 32 32 32 32 32 32 32 31 32 32
-2006 0.618 0.907 0.767 0.580 0.523 0.869 0.801 0.736 0.772 0.950 0.874 1.303 0.808
-2006 31 32 32 32 32 32 33 32 32 32 33 31
-2007 1.670 0.886 1.009 1.109 0.790 0.594 0.633 0.792 0.832 0.987 0.784 0.804 0.907
-2007 32 32 32 32 32 33 33 32 33 32 32 32
-2008 0.418 0.531 1.194 0.572 0.495 0.624 0.741 0.589 0.576 1.001 1.057 0.671 0.706
-2008 32 32 32 33 33 32 32 33 33 33 33 32
-2009 0.778 0.790 0.639 0.826 0.598 0.665 0.596 0.867 0.965 0.718 0.853 0.484 0.731
-2009 32 33 32 31 32 32 33 32 32 32 32 32
-2010 0.767 0.804 1.087 1.040 0.917 0.982 0.965 0.859 0.730 0.874 1.170 0.454 0.887
-2010 33 32 33 32 32 32 33 33 33 33 33 33
-2011 0.482 0.398 0.596 0.836 0.536 0.777 0.778 0.775 0.851 0.933 0.621 0.832 0.701
-2011 31 32 31 31 31 31 31 31 31 31 31 31
-2012 0.589 0.352 0.679 1.019 0.953 0.934 0.799 0.884 0.896 0.834 0.838 0.290 0.756
-2012 31 30 29 29 29 29 29 29 30 29 29 29
-2013 0.919 0.938 0.663 0.588 0.768 0.818 0.680 0.706 0.771 0.810 0.969 0.786 0.784
-2013 28 28 28 28 29 29 28 28 28 28 29 28
-2014 0.930 0.421 0.921 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.757
-2014 27 28 28 0 0 0 0 0 0 0 0 0
diff --git a/examples/temperature/temperature.jl b/examples/temperature/temperature.jl
deleted file mode 100644
index e041e46..0000000
--- a/examples/temperature/temperature.jl
+++ /dev/null
@@ -1,45 +0,0 @@
-"""
-Reads a temperature data file as taken from
-http://www.cru.uea.ac.uk/cru/data/temperature/
-"""
-import PyPlot.plt
-using LinearLeastSquares
-
-data = readdlm("CRUTEM4v-gl.dat", ' ');
-
-# We only want to extract the odd lines (1-indexed)
-data = vcat([data[i, :] for i in 1 : 2 : size(data, 1)]...);
-data = float(data);
-years = data[:, 1];
-temperatures = vec(data[:, 2 : end - 1]');
-
-num_samples = size(temperatures, 1)
-
-yearly_tread = Variable(num_samples);
-a = Variable();
-b = Variable();
-
-
-equality_constraints = EqConstraint[];
-for i in 13 : num_samples
-  push!(equality_constraints, yearly_tread[i] == yearly_tread[i - 12])
-end
-
-t = [0 : num_samples - 1];
-t2 = t .^ 2;
-
-
-smoothing = 0.01;
-objective = sum_squares(yearly_tread + t * a + t2 * b - temperatures);
-objective += smoothing * sum_squares(yearly_tread[1 : num_samples - 1] - yearly_tread[2 : num_samples]);
-
-optval = minimize!(objective, equality_constraints);
-
-plt.figure(0)
-plt.plot(temperatures)
-plt.plot(yearly_tread.value + t*a.value + t2*b.value, color="r")
-plt.show()
-
-# plt.figure(1)
-# plt.plot(yearly_tread.value + t*b.value + t2*a.value - temperatures)
-# plt.show()
diff --git a/examples/time_series/ar_fit.png b/examples/time_series/ar_fit.png
new file mode 100644
index 0000000..5fc81fc
Binary files /dev/null and b/examples/time_series/ar_fit.png differ
diff --git a/examples/time_series/melbourne.png b/examples/time_series/melbourne.png
new file mode 100644
index 0000000..c9ad5dd
Binary files /dev/null and b/examples/time_series/melbourne.png differ
diff --git a/examples/time_series/residuals.png b/examples/time_series/residuals.png
new file mode 100644
index 0000000..4e98d7a
Binary files /dev/null and b/examples/time_series/residuals.png differ
diff --git a/examples/time_series/time_series.jl b/examples/time_series/time_series.jl
index de50957..c3d49a8 100644
--- a/examples/time_series/time_series.jl
+++ b/examples/time_series/time_series.jl
@@ -1,34 +1,39 @@
 using LinearLeastSquares
-import PyPlot.plt
-temps = readdlm("melbourne_temps.txt", ',');
-plt.figure(0)
-plt.plot(temps, color="b")
-plt.title("Melbourne Daily Temperature")
-n = size(temps, 1);
-plt.xlim([0, n])
+using Gadfly
 
-yearly = Variable(n)
+temps = readdlm("melbourne_temps.txt", ',')
+n = size(temps)[1]
+p = plot(
+  x=1:1500, y=temps[1:1500], Geom.line,
+  Theme(panel_fill=color("white"))
+)
+# draw(PNG("melbourne.png", 16cm, 12cm), p)
 
+yearly = Variable(n)
 eq_constraints = EqConstraint[]
 for i in 365 + 1 : n
-  eq_constraints += yearly[i] == yearly[i - 365];
+  eq_constraints += yearly[i] == yearly[i - 365]
 end
 
-smoothing = 100;
-smooth_objective = sum_squares(yearly[1 : n - 1] - yearly[2 : n]);
-optval = minimize!(sum_squares(temps - yearly) + smoothing * smooth_objective, eq_constraints);
-residuals = temps - evaluate(yearly);
+smoothing = 100
+smooth_objective = sum_squares(yearly[1 : n - 1] - yearly[2 : n])
+optval = minimize!(sum_squares(temps - yearly) + smoothing * smooth_objective, eq_constraints)
+residuals = temps - evaluate(yearly)
 
 # Plot smooth fit
-plt.figure(1)
-plt.plot(temps)
-plt.plot(evaluate(yearly), color="r")
-plt.title("Smooth Fit of Data")
-plt.xlim([0, n])
+p = plot(
+  layer(x=1:1500, y=evaluate(yearly)[1:1500], Geom.line, Theme(default_color=color("red"), line_width=2px)),
+  layer(x=1:1500, y=temps[1:1500], Geom.line),
+  Theme(panel_fill=color("white"))
+)
+# draw(PNG("yearly_fit.png", 16cm, 12cm), p)
 
 # Plot residuals for a few days
-plt.figure(2)
-plt.plot(residuals[1:100], color="g")
+p = plot(
+  x=1:100, y=residuals[1:100], Geom.line,
+  Theme(default_color=color("green"), panel_fill=color("white"))
+)
+# draw(PNG("residuals.png", 16cm, 12cm), p)
 
 # Generate the residuals matrix
 ar_len = 5
@@ -41,19 +46,25 @@ end
 ar_coef = Variable(ar_len)
 optval2 = minimize!(sum_squares(residuals_mat * ar_coef - residuals[ar_len + 1 : end]))
 
-# plot autoregressive fit of daily fluctuations for first few days
-plt.figure(3)
-plt.plot(residuals[ar_len + 1 : ar_len + 100], color="g")
-plt.plot(residuals_mat[1:100, :] * evaluate(ar_coef), color="r")
-plt.title("Autoregressive Fit of Residuals")
+# plot autoregressive fit of daily fluctuations for a few days
+ar_range = 1:145
+day_range = ar_range + ar_len
+p = plot(
+  layer(x=day_range, y=residuals[day_range], Geom.line, Theme(default_color=color("green"))),
+  layer(x=day_range, y=residuals_mat[ar_range, :] * evaluate(ar_coef), Geom.line, Theme(default_color=color("red"))),
+  Theme(panel_fill=color("white"))
+)
+# draw(PNG("ar_fit.png", 16cm, 12cm), p)
+
 
-# plot final fit of data
-plt.figure(4)
-plt.plot(temps)
 total_estimate = evaluate(yearly)
 total_estimate[ar_len + 1 : end] += residuals_mat * evaluate(ar_coef)
-plt.plot(total_estimate, color="r", alpha=0.5)
-plt.title("Total Fit of Data")
-plt.xlim([0, n])
 
-plt.show()
+# plot final fit of data
+p = plot(
+  layer(x=1:1500, y=total_estimate[1:1500], Geom.line, Theme(default_color=color("red"))),
+  layer(x=1:1500, y=temps[1:1500], Geom.line),
+  Theme(panel_fill=color("white"))
+)
+# draw(PNG("total_fit.png", 16cm, 12cm), p)
+
diff --git a/examples/time_series/total_fit.png b/examples/time_series/total_fit.png
new file mode 100644
index 0000000..030ed31
Binary files /dev/null and b/examples/time_series/total_fit.png differ
diff --git a/examples/time_series/yearly_fit.png b/examples/time_series/yearly_fit.png
new file mode 100644
index 0000000..a763dfe
Binary files /dev/null and b/examples/time_series/yearly_fit.png differ
diff --git a/examples/tomography/tomography.jl b/examples/tomography/tomography.jl
index f29a408..6b5e9c3 100644
--- a/examples/tomography/tomography.jl
+++ b/examples/tomography/tomography.jl
@@ -1,9 +1,10 @@
 using LinearLeastSquares
+using Gadfly
 
-line_mat_x = readdlm("tux_sparse_x.txt");
-line_mat_y = readdlm("tux_sparse_y.txt");
-line_mat_val = readdlm("tux_sparse_val.txt");
-line_vals = readdlm("tux_sparse_lines.txt");
+line_mat_x = readdlm("tux_sparse_x.txt")
+line_mat_y = readdlm("tux_sparse_y.txt")
+line_mat_val = readdlm("tux_sparse_val.txt")
+line_vals = readdlm("tux_sparse_lines.txt")
 
 # Form the sparse matrix from the data
 # Image is 50 x 50
@@ -11,21 +12,33 @@ img_size = 50
 # The number of pixels in the image
 num_pixels = img_size * img_size
 
-line_mat = spzeros(3300, num_pixels);
+line_mat = spzeros(3300, num_pixels)
 
 num_vals = length(line_mat_val)
 
 for i in 1:num_vals
-  x = int(line_mat_x[i]);
-  y = int(line_mat_y[i]);
-  line_mat[x + 1, y + 1] = line_mat_val[i];
+  x = int(line_mat_x[i])
+  y = int(line_mat_y[i])
+  line_mat[x + 1, y + 1] = line_mat_val[i]
 end
 
 x = Variable(num_pixels)
-objective = sum_squares(line_mat * x - line_vals);
-optval = minimize!(objective);
+objective = sum_squares(line_mat * x - line_vals)
+optval = minimize!(objective)
 
-import PyPlot.plt
-import PyPlot.cm
-using PyPlot
-plt.imshow(reshape(evaluate(x), img_size,img_size), cmap = get_cmaps()[29])
+rows = zeros(img_size*img_size)
+cols = zeros(img_size*img_size)
+for i = 1:img_size
+  for j = 1:img_size
+    rows[(i-1)*img_size + j] = i
+    cols[(i-1)*img_size + j] = img_size + 1 - j
+  end
+end
+
+p = plot(
+  x=rows, y=cols, color=reshape(evaluate(x), img_size, img_size), Geom.rectbin,
+  Scale.ContinuousColorScale(Scale.lab_gradient(color("black"), color("white")))
+)
+draw(PNG("tomography.png", 16cm, 14cm), p)
+
+# plt.imshow(reshape(evaluate(x), img_size, img_size), cmap = get_cmaps()[29])
diff --git a/examples/tomography/tomography.png b/examples/tomography/tomography.png
new file mode 100644
index 0000000..6aa45f9
Binary files /dev/null and b/examples/tomography/tomography.png differ
diff --git a/src/atoms/add_subtract.jl b/src/atoms/add_subtract.jl
index 9dbb79b..ad022c6 100644
--- a/src/atoms/add_subtract.jl
+++ b/src/atoms/add_subtract.jl
@@ -12,7 +12,7 @@ function -(x::AffineExpr)
     vars_to_coeffs_map[v] = -c
   end
   this = AffineExpr(:-, (x,), vars_to_coeffs_map, -x.constant, x.size)
-  this.evaluate = ()-> -this.value;
+  this.evaluate = ()->(-x.evaluate())
   return this
 end
 
@@ -78,8 +78,10 @@ end
 
 function +(x::SumSquaresExpr, y::SumSquaresExpr)
   affines = copy(x.affines)
+  multipliers = copy(x.multipliers)
   append!(affines, y.affines)
-  this = SumSquaresExpr(:+, affines)
+  append!(multipliers, y.multipliers)
+  this = SumSquaresExpr(:+, affines, multipliers)
   return this
 end
 
diff --git a/src/atoms/diag.jl b/src/atoms/diag.jl
index 3677011..9cf21dd 100644
--- a/src/atoms/diag.jl
+++ b/src/atoms/diag.jl
@@ -14,10 +14,16 @@ function diag(x::AffineExpr, num::Int64)
   end
   len = min(x.size[1] - start_row + 1, x.size[2] - start_col + 1)
   start_ind = (start_col - 1) * x.size[1] + start_row
-  indexer = Constant(spzeros(len, x.size[1] * x.size[2]))
-  for i = 1 : len
-    indexer.value[i, start_ind + (i - 1) * (x.size[1] + 1)] = 1
-  end
+
+  indexer = Constant(
+    sparse(
+      1 : len,
+      start_ind : x.size[1] + 1 : start_ind + (len - 1) * (x.size[1] + 1),
+      1.0,
+      len,
+      x.size[1] * x.size[2]
+    )
+  )
 
   vars_to_coeffs_map = Dict{Uint64, Constant}()
   for (v, c) in x.vars_to_coeffs_map
@@ -35,10 +41,16 @@ function diagm(x::AffineExpr)
   if x.size[2] != 1
     error("Can only call diagm on column vectors")
   end
-  indexer = Constant(spzeros(x.size[1] * x.size[1], x.size[1]))
-  for i = 1 : x.size[1]
-    indexer.value[1 + (i - 1) * (x.size[1] + 1), i] = 1
-  end
+
+  indexer = Constant(
+    sparse(
+      1 : x.size[1] + 1 : 1 + (x.size[1] - 1) * (x.size[1] + 1),
+      1 : x.size[1],
+      1.0,
+      x.size[1] * x.size[1],
+      x.size[1]
+    )
+  )
 
   vars_to_coeffs_map = Dict{Uint64, Constant}()
   for (v, c) in x.vars_to_coeffs_map
diff --git a/src/atoms/index.jl b/src/atoms/index.jl
index b5e0eb9..57ed627 100644
--- a/src/atoms/index.jl
+++ b/src/atoms/index.jl
@@ -6,19 +6,17 @@ getindex{T <: Real}(x::Constant, inds::AbstractArray{T, 1}) = Constant(getindex(
 getindex{T <: Real}(x::Constant, rows::AbstractArray{T, 1}, cols::AbstractArray{T, 1}) = Constant(getindex(x.value, rows, cols))
 
 function getindex{T <: Real}(x::AffineExpr, inds::AbstractArray{T, 1})
-  indexer = Constant(spzeros(length(inds), x.size[1] * x.size[2]))
-  k = 1
-  for i in inds
-    indexer.value[k, i] = 1
-    k += 1
-  end
+  # number of rows/cols in the coefficient for x in our canonical form
+  num_rows_coeff = length(inds)
+  num_cols_coeff = x.size[1] * x.size[2]
+  indexer = Constant(sparse(1:length(inds), inds, 1.0, num_rows_coeff, num_cols_coeff))
 
   vars_to_coeffs_map = Dict{Uint64, Constant}()
   for (v, c) in x.vars_to_coeffs_map
     vars_to_coeffs_map[v] = indexer * c
   end
   constant = indexer * x.constant
-  this = AffineExpr(:index, (x,), vars_to_coeffs_map, constant, (length(inds), 1))
+  this = AffineExpr(:index, (x,), vars_to_coeffs_map, constant, (num_rows_coeff, 1))
   this.evaluate = ()->x.evaluate()[inds]
   return this
 end
@@ -28,18 +26,17 @@ function getindex{T <: Real}(x::AffineExpr, rows::AbstractArray{T, 1}, cols::Abs
   num_rows_coeff = length(rows) * length(cols)
   num_cols_coeff = x.size[1] * x.size[2]
 
-  indexer = Constant(spzeros(num_rows_coeff, num_cols_coeff))
-
   # Create the indexing matrix such that indexer * vec(x) = vec(x[rows, cols])
+  J = Array(Int64, num_rows_coeff)
   k = 1
   num_rows = x.size[1]
   for c in cols
     for r in rows
-      idx = num_rows * (convert(Int64, c) - 1) + convert(Int64, r)
-      indexer.value[k, idx] = 1
+      J[k] = num_rows * (convert(Int64, c) - 1) + convert(Int64, r)
       k += 1
     end
   end
+  indexer = Constant(sparse(1:num_rows_coeff, J, 1.0, num_rows_coeff, num_cols_coeff))
 
   vars_to_coeffs_map = Dict{Uint64, Constant}()
   for (v, c) in x.vars_to_coeffs_map
diff --git a/src/atoms/multiply_divide.jl b/src/atoms/multiply_divide.jl
index c49d2a3..8c196c9 100644
--- a/src/atoms/multiply_divide.jl
+++ b/src/atoms/multiply_divide.jl
@@ -92,38 +92,40 @@ end
 
 # Only support scalar division
 function /(x::AffineExpr, y::Constant)
-  if y.size == (1, 1)
-    return x ./ y
+  if y.value == 0
+    error("Cannot divide an affine expression by 0")
   else
-    error("Cannot divide by an expression whose size is not 1 by 1")
+    return x ./ y
   end
 end
 
-./(x::AffineExpr, y::Numeric) = ./(x, convert(Constant, y))
-/(x::AffineExpr, y::Numeric) = /(x, convert(Constant, y))
+/(x::AffineExpr, y::Number) = /(x, convert(Constant, y))
 
 
 ## Sum of squares expressions
 
-function *(x::Constant, y::SumSquaresExpr)
-  if x.size != (1, 1) || x.value < 0
+function *(x::Number, y::SumSquaresExpr)
+  try
+    x = convert(Float64, x)
+  catch
+    error("Sum Squares expressions can only be multiplied by real numbers")
+  end
+  if x < 0
     error("Sum Squares expressions can only be multiplied by nonegative scalars")
   end
-  x_sqrt = Constant(sqrt(x.value))
-  affines = [x_sqrt * affine for affine in y.affines]
-  this = SumSquaresExpr(:*, affines)
-  return this
+  return SumSquaresExpr(:*, y.affines, x * y.multipliers)
 end
 
-*(x::SumSquaresExpr, y::Constant) = *(y, x)
-*(x::SumSquaresExpr, y::Numeric) = *(x, convert(Constant, y))
-*(x::Numeric, y::SumSquaresExpr) = *(convert(Constant, x), y)
+*(x::SumSquaresExpr, y::Number) = *(y, x)
 
-function /(x::SumSquaresExpr, y::Constant)
-  if y.size != (1, 1) || all(y.value .<= 0)
+function /(x::SumSquaresExpr, y::Number)
+  try
+    y = convert(Float64, y)
+  catch
+    error("Sum Squares expressions can only be divided by real numbers")
+  end
+  if y <= 0
     error("Sum Squares expressions can only be divided by positive scalars")
   end
-  return Constant(1 ./ y.value) * x
+  return SumSquaresExpr(:/, x.affines, x.multipliers / y)
 end
-
-/(x::SumSquaresExpr, y::Numeric) = /(x, convert(Constant, y))
diff --git a/src/atoms/stack.jl b/src/atoms/stack.jl
index ea7d4a2..d545590 100644
--- a/src/atoms/stack.jl
+++ b/src/atoms/stack.jl
@@ -79,5 +79,11 @@ function hcat(args::AffineExpr...)
 end
 
 function hvcat(rows::(Int64...), args::AffineExpr...)
-  error("hvcat not yet implemented")
+  row_exprs = AffineExpr[]
+  index = 0
+  for row_size in rows
+    push!(row_exprs, hcat(args[index + 1 : index + row_size]...))
+    index += row_size
+  end
+  return vcat(row_exprs...)
 end
diff --git a/src/atoms/transpose.jl b/src/atoms/transpose.jl
index 7b026f8..7f7963b 100644
--- a/src/atoms/transpose.jl
+++ b/src/atoms/transpose.jl
@@ -6,19 +6,22 @@ ctranspose(x::Constant) = transpose(x)
 
 function transpose(x::AffineExpr)
   vec_sz = x.size[1] * x.size[2]
-  perm_matrix = spzeros(vec_sz, vec_sz)
   num_rows = x.size[1]
   num_cols = x.size[2]
 
+  I = Array(Int64, vec_sz)
+  J = Array(Int64, vec_sz)
+
+  k = 1
   for r = 1:num_rows
     for c = 1:num_cols
-      i = (c - 1) * num_rows + r
-      j = (r - 1) * num_cols + c
-      perm_matrix[i, j] = 1.0
+      J[k] = (c - 1) * num_rows + r
+      I[k] = (r - 1) * num_cols + c
+      k += 1
     end
   end
 
-  perm_constant = Constant(perm_matrix)
+  perm_constant = Constant(sparse(I, J, 1.0))
 
   vars_to_coeffs_map = Dict{Uint64, Constant}()
   for (v, c) in x.vars_to_coeffs_map
@@ -26,7 +29,7 @@ function transpose(x::AffineExpr)
   end
   constant = perm_constant * x.constant
 
-  this = AffineExpr(:transpose, (x,), vars_to_coeffs_map, constant, (x.size[2], x.size[1]))
+  this = AffineExpr(:transpose, (x,), vars_to_coeffs_map, constant, (num_cols, num_rows))
   this.evaluate = ()->x.evaluate()'
   return this
 end
diff --git a/src/solve.jl b/src/solve.jl
index 81b1517..e0b8c5a 100644
--- a/src/solve.jl
+++ b/src/solve.jl
@@ -1,4 +1,4 @@
-export minimize!, satisfy!
+export minimize!, solve!
 
 # Eventually, this can take in arguments to specify solving method.
 # For now, backslash is the only supported method.
@@ -6,9 +6,8 @@ function solve!(problem_type::Symbol, objective::SumSquaresExpr, constraints::Ar
   p = Problem(problem_type, objective, constraints)
   # return backslash_solve!(p)
   backslash_solve!(p)
-  if p.status == "infeasible"
-    println("Problem was infeasible")
-    return nothing
+  if p.status == "KKT singular"
+    error("KKT system is singular. No unique solution")
   elseif p.status == "solved"
     return p.optval
   else
@@ -19,8 +18,9 @@ end
 
 minimize!(objective::SumSquaresExpr, constraints::Array{EqConstraint}) = solve!(:minimize, objective, constraints)
 minimize!(objective::SumSquaresExpr, constraint::EqConstraint) = minimize!(objective, [constraint])
+minimize!(objective::SumSquaresExpr, constraints::EqConstraint...) = minimize!(objective, [constraints...])
 minimize!(objective::SumSquaresExpr) = minimize!(objective, EqConstraint[])
 
-solve!(constraints::Array{EqConstraint}) = solve!(:satisfy, SumSquaresExpr(:empty, AffineExpr[]), constraints)
-solve!(constraint::EqConstraint) = satisfy!([constraint])
-solve!(constraints::EqConstraint...) = satisfy!([constraints...])
+solve!(constraints::Array{EqConstraint}) = solve!(:solve, SumSquaresExpr(:empty, AffineExpr[], Float64[]), constraints)
+solve!(constraint::EqConstraint) = solve!([constraint])
+solve!(constraints::EqConstraint...) = solve!([constraints...])
diff --git a/src/solvers/backslash_solver.jl b/src/solvers/backslash_solver.jl
index 999482f..96c1d98 100644
--- a/src/solvers/backslash_solver.jl
+++ b/src/solvers/backslash_solver.jl
@@ -1,4 +1,4 @@
-export backslash_solve!
+export backslash_solve!, build_kkt_system
 
 function reset_value_and_add_vars!(x::AffineOrConstant, unique_vars_map::Dict{Uint64, AffineExpr})
   if x.head != :constant
@@ -78,11 +78,14 @@ function build_kkt_system(A, b, C, d)
     constant[1 : size(A, 2)] = -2 * A' * b
     if size(C, 1) != 0
       coefficient[1 : size(C, 2), size(A, 2) + 1 : end] = C'
+      coefficient[size(A, 2) + 1 : end, 1 : size(C, 2)] = C
+      constant[size(A, 2) + 1 : end] = -d
     end
-  end
-  if size(C, 1) != 0
-    coefficient[size(A, 2) + 1 : end, 1 : size(C, 2)] = C
-    constant[size(A, 2) + 1 : end] = -d
+  elseif size(C, 1) != 0
+    coefficient[:, 1 : size(C, 2)] = C
+    constant = -d
+  else
+    error("KKT system should not be empty")
   end
 
   return coefficient, constant
@@ -104,19 +107,24 @@ function backslash_solve!(p::Problem)
     push!(canon_forms, constraint.canon_form)
   end
   C, d = coalesce_affine_exprs(vars_to_ranges_map, num_vars, canon_forms)
-  # return A, b, C, d
-  coefficient, constant = build_kkt_system(A, b, C, d)
   solution = nothing
-  try
-    solution = coefficient\full(constant)
-  catch
-    println("KKT system is singular")
+  # return A, b, C, d
+  if size(A, 1) > 0 || size(C, 1) > 0
+    coefficient, constant = build_kkt_system(A, b, C, d)
+    try
+      # sparse solver doesnt error out some 2x2 singular KKT systems
+      solution = coefficient\full(constant)
+    end
   end
   if solution == nothing
-    p.status = "infeasible"
+    p.status = "KKT singular"
   else
     p.status = "solved"
-    p.optval = norm(A*solution[1:num_vars] + b)^2
+    if size(A, 1) > 0
+      p.optval = sum((A*solution[1:num_vars] + b).^2)
+    else
+      p.optval = 0.0
+    end
     populate_vars!(unique_vars_map, vars_to_ranges_map, solution)
   end
   return p.optval
diff --git a/src/types/expressions.jl b/src/types/expressions.jl
index 8285ff8..953b59c 100644
--- a/src/types/expressions.jl
+++ b/src/types/expressions.jl
@@ -12,17 +12,18 @@ Numeric = Union(Number, AbstractArray)
 ValueOrNothing = Union(Value, Nothing)
 
 abstract AbstractExpr
+abstract AffineOrConstant <: AbstractExpr
 
 function evaluate(x::AbstractExpr)
   if x.size == (1, 1)
     return (x.evaluate())[1]
+  elseif x.size[2] == 1
+    return vec(full(x.evaluate()))
   else
     return full(x.evaluate())
   end
 end
 
-abstract AffineOrConstant <: AbstractExpr
-
 type Constant <: AffineOrConstant
   head::Symbol
   value::ValueOrNothing
@@ -72,6 +73,9 @@ function AffineConstant(value::Value)
 end
 
 function Variable(size::(Int64, Int64))
+  if (size[1] < 1 || size[2] < 1)
+    error("invalid variable size")
+  end
   vec_sz = size[1]*size[2]
   this = AffineExpr(:variable, (), Dict{Uint64, Constant}(), Constant(spzeros(vec_sz, 1)), size)
   this.vars_to_coeffs_map[this.uid] = Constant(speye(vec_sz))
@@ -95,18 +99,19 @@ type SumSquaresExpr <: AbstractExpr
   head::Symbol
   value::ValueOrNothing
   affines::Array{AffineExpr}
+  multipliers::Array{Float64}
   uid::Uint64
   evaluate::Function
 
-  function SumSquaresExpr(head::Symbol, affines::Array{AffineExpr})
-    this = new(head, nothing, affines)
+  function SumSquaresExpr(head::Symbol, affines::Array{AffineExpr}, multipliers::Array{Float64})
+    this = new(head, nothing, affines, multipliers)
     this.uid = object_id(this)
     this.evaluate = ()->begin
-      sum = 0
-      for affine in affines
-        sum += norm(vec(affine.evaluate()))^2
+      value = 0
+      for i in 1:length(this.affines)
+        value += multipliers[i] * sum((this.affines[i].evaluate()).^2)
       end
-      return sum
+      return value
     end
     return this
   end
@@ -115,7 +120,11 @@ end
 function sum_squares(affine::AffineExpr)
   affines = AffineExpr[]
   push!(affines, affine)
-  this = SumSquaresExpr(:sum_squares, affines)
+  this = SumSquaresExpr(:sum_squares, affines, ones(length(affines)))
+end
+
+function evaluate(x::SumSquaresExpr)
+  return (x.evaluate())[1]
 end
 
 endof(x::AffineOrConstant) = x.size[1]*x.size[2]
diff --git a/test/runtests.jl b/test/runtests.jl
new file mode 100644
index 0000000..ab231ce
--- /dev/null
+++ b/test/runtests.jl
@@ -0,0 +1 @@
+include("unit_tests/unit_tests.jl")
diff --git a/test/unit_tests/add_subtract.jl b/test/unit_tests/add_subtract.jl
new file mode 100644
index 0000000..92a7626
--- /dev/null
+++ b/test/unit_tests/add_subtract.jl
@@ -0,0 +1,19 @@
+a = create_var(1, 1)
+b = eye(3)
+c = a + b
+test_expr(c, (a,))
+
+p = b + c
+test_expr(p, (a,))
+
+x = create_var(3, 4)
+d = a - x
+test_expr(d, (a, x))
+
+f = x + a
+test_expr(f, (a, x))
+
+g = x - a
+test_expr(g, (a, x))
+
+info("All add/subtract tests passed")
diff --git a/test/unit_tests/compound.jl b/test/unit_tests/compound.jl
new file mode 100644
index 0000000..db97015
--- /dev/null
+++ b/test/unit_tests/compound.jl
@@ -0,0 +1,22 @@
+# tests for sum, mean, and var, which implemented by calling other atoms
+
+x = create_var(5, 3)
+s = sum(x)
+m = mean(x)
+test_expr(s, (x,))
+test_expr(m, (x,))
+
+t = sum(x, 1)
+n = mean(x, 1)
+test_expr(t, (x,))
+test_expr(n, (x,))
+
+u = sum(x, 2)
+p = mean(x, 2)
+test_expr(u, (x,))
+test_expr(p, (x,))
+
+v = var(x)
+test_expr(v)
+
+info("All sum/mean/var tests passed")
diff --git a/test/unit_tests/constraints.jl b/test/unit_tests/constraints.jl
new file mode 100644
index 0000000..4ff464c
--- /dev/null
+++ b/test/unit_tests/constraints.jl
@@ -0,0 +1,9 @@
+x = Variable(4, 3)
+y = Variable()
+
+x + y == 3
+3 == y
+rand(4, 3) == x - 4*y
+sum(y, 1) == rand(1, 3)
+
+info("All equality constraints tests passed")
diff --git a/test/unit_tests/diag.jl b/test/unit_tests/diag.jl
new file mode 100644
index 0000000..4a52e8c
--- /dev/null
+++ b/test/unit_tests/diag.jl
@@ -0,0 +1,20 @@
+# diagm test
+x = create_var(4, 1)
+d = diagm(x)
+test_expr(d, (x,))
+
+# main diagonal
+x = create_var(3, 4)
+d = diag(x, 0)
+test_expr(d, (x,))
+
+# off diagonals
+x = create_var(3, 4)
+d = diag(x, 2)
+test_expr(d, (x,))
+
+x = create_var(3, 4)
+d = diag(x, -2)
+test_expr(d, (x,))
+
+info("All diag tests passed")
diff --git a/test/unit_tests/helpers.jl b/test/unit_tests/helpers.jl
new file mode 100644
index 0000000..8f9d707
--- /dev/null
+++ b/test/unit_tests/helpers.jl
@@ -0,0 +1,46 @@
+export eval_internals, set_value!, test_expr, create_var
+
+function eval_internals(x::AffineExpr, vars::Tuple)
+  (rows, cols) = x.size
+  value = zeros(rows, cols)
+  for var in vars
+    coeff = x.vars_to_coeffs_map[object_id(var)]
+    value += reshape(coeff.value * vec(var.value), rows, cols)
+  end
+  value += reshape(x.constant.value, rows, cols)
+  return value
+end
+
+function eval_internals(x::SumSquaresExpr)
+  value = 0
+  for i in 1:length(x.affines)
+    value += x.multipliers[i]*vecnorm(evaluate(x.affines[i]))^2
+  end
+  return value
+end
+
+function set_value!(x::AffineExpr, value::Number)
+  x.value = convert(Float64, value)
+end
+
+function set_value!(x::AffineExpr, value::AbstractArray)
+  x.value = convert(SparseMatrixCSC{Float64, Int64}, sparse(value))
+end
+
+function test_expr(x::SumSquaresExpr)
+  val1 = evaluate(x)
+  val2 = eval_internals(x)
+  @assert abs(val1 .- val2) <= 0.0001
+end
+
+function test_expr(x::AffineExpr, vars::Tuple)
+  val1 = eval_internals(x, vars)
+  val2 = evaluate(x)
+  @assert all(abs(val1 .- val2) .<= 0.0001)
+end
+
+function create_var(m::Int64, n::Int64)
+  x = Variable(m, n)
+  set_value!(x, rand(m, n))
+  return x
+end
diff --git a/test/unit_tests/index.jl b/test/unit_tests/index.jl
new file mode 100644
index 0000000..6cc4bcc
--- /dev/null
+++ b/test/unit_tests/index.jl
@@ -0,0 +1,8 @@
+# indexing tests
+x = create_var(4, 6)
+y = x[1:3, 5:end]
+z = x[:]
+test_expr(y, (x,))
+test_expr(z, (x,))
+
+info("All index tests passed")
diff --git a/test/unit_tests/multiply_divide.jl b/test/unit_tests/multiply_divide.jl
new file mode 100644
index 0000000..3832aed
--- /dev/null
+++ b/test/unit_tests/multiply_divide.jl
@@ -0,0 +1,19 @@
+a = create_var(1, 1)
+b = eye(3)
+c = a * b
+test_expr(c, (a,))
+
+d = create_var(5, 3)
+f = a * 3.1 + 5.3 * d
+test_expr(f, (a, d))
+
+g = rand(4, 5) * d
+test_expr(g, (d,))
+
+h = d * rand(3, 4)
+test_expr(h, (d,))
+
+j = d / 4
+test_expr(j, (d,))
+
+info("All multiply/divide tests passed")
diff --git a/test/unit_tests/problems.jl b/test/unit_tests/problems.jl
new file mode 100644
index 0000000..57bfce1
--- /dev/null
+++ b/test/unit_tests/problems.jl
@@ -0,0 +1,47 @@
+TOLERANCE = 0.0001
+
+x = Variable()
+solve!(x == 3)
+@assert evaluate(x) == 3
+
+y = Variable(3, 1)
+A = rand(3, 3)
+solve!(x == 3, A*y == 4)
+@assert evaluate(x) == 3
+@assert all(abs(A*evaluate(y) - 4) .<= TOLERANCE)
+
+solve!([x == 5, y == 7])
+@assert evaluate(x) == 5
+@assert all(evaluate(y) .== 7)
+
+x = Variable(3)
+y = sum_squares(x)
+optval = minimize!(y)
+@assert abs(optval - 0) < TOLERANCE
+@assert all(abs(evaluate(x)) .<= TOLERANCE)
+
+x = Variable(3)
+y = sum_squares(x)
+optval = minimize!(y, x == 3)
+@assert abs(optval - 27) < TOLERANCE
+@assert all(abs(evaluate(x) .- 3) .<= TOLERANCE)
+
+x = Variable(2)
+A = [1 0; 1 0]
+y = sum_squares(x)
+@test_throws ErrorException optval = minimize!(y, A * x == [1; 2])
+
+A = randn(5, 5)
+x_real = randn(5, 5)
+x = Variable(5, 5)
+y = A * x
+b = A * x_real
+constraints = EqConstraint[]
+for i = 1:5
+  constraints += y[:, i] == b[:, i]
+end
+optval = minimize!(sum_squares(x), constraints)
+@assert abs(sum(x_real.^2) - optval) <= TOLERANCE
+@assert all(abs(evaluate(x) .- x_real) .<= TOLERANCE)
+
+info("All solve/minimize tests passed")
diff --git a/test/unit_tests/repmat.jl b/test/unit_tests/repmat.jl
new file mode 100644
index 0000000..a8d8df1
--- /dev/null
+++ b/test/unit_tests/repmat.jl
@@ -0,0 +1,19 @@
+x = create_var(3, 2)
+r = repmat(x, 4, 5)
+test_expr(r, (x,))
+
+x = create_var(1, 1)
+r = repmat(x, 1, 1)
+test_expr(r, (x,))
+
+x = create_var(3, 3)
+y = x + [1 2 3; 4 5 6; 7 8 9]
+r = repmat(y, 1, 3)
+test_expr(r, (x,))
+
+x = create_var(3, 3)
+y = x + [1 2 3; 4 5 6; 7 8 9]
+r = repmat(y, 4, 1)
+test_expr(r, (x,))
+
+info("All repmat tests passed")
diff --git a/test/unit_tests/reshape.jl b/test/unit_tests/reshape.jl
new file mode 100644
index 0000000..695503f
--- /dev/null
+++ b/test/unit_tests/reshape.jl
@@ -0,0 +1,10 @@
+# reshape test
+x = create_var(6, 4)
+y = reshape(x, 8, 3)
+z = reshape(x, 4, 6)
+v = vec(x)
+test_expr(y, (x,))
+test_expr(z, (x,))
+test_expr(v, (x,))
+
+info("All reshape tests passed")
diff --git a/test/unit_tests/stack.jl b/test/unit_tests/stack.jl
new file mode 100644
index 0000000..9fb2857
--- /dev/null
+++ b/test/unit_tests/stack.jl
@@ -0,0 +1,21 @@
+# hcat test
+x = create_var(3, 1)
+y = create_var(3, 2)
+z = create_var(3, 3)
+h = [x y z]
+test_expr(h, (x, y, z))
+
+# vcat test
+x = create_var(1, 3)
+y = create_var(2, 3)
+z = create_var(3, 3)
+v = [x; y; z]
+test_expr(v, (x, y, z))
+v = [x, z, y]
+test_expr(v, (x, y, z))
+
+# hvcat test
+hv = [x x; y y; z z]
+test_expr(hv, (x, y, z))
+
+info("All stack tests passed")
diff --git a/test/unit_tests/sum_squares.jl b/test/unit_tests/sum_squares.jl
new file mode 100644
index 0000000..4b87364
--- /dev/null
+++ b/test/unit_tests/sum_squares.jl
@@ -0,0 +1,15 @@
+x = create_var(1, 1)
+y = create_var(3, 1)
+s = sum_squares(x) + sum_squares(y)
+test_expr(s)
+
+s = 4 * s
+test_expr(s)
+
+s = s * 2
+test_expr(s)
+
+s = s / 4
+test_expr(s)
+
+info("All sum squares tests passed")
diff --git a/test/unit_tests/transpose.jl b/test/unit_tests/transpose.jl
new file mode 100644
index 0000000..de56572
--- /dev/null
+++ b/test/unit_tests/transpose.jl
@@ -0,0 +1,9 @@
+x = create_var(3, 4)
+y = x'
+test_expr(y, (x,))
+
+x = create_var(10, 1)
+y = x''
+test_expr(y, (x,))
+
+info("All transpose tests passed")
diff --git a/tests/unit_tests/unit_tests.jl b/test/unit_tests/unit_tests.jl
similarity index 56%
rename from tests/unit_tests/unit_tests.jl
rename to test/unit_tests/unit_tests.jl
index 0cc2715..4bae10b 100644
--- a/tests/unit_tests/unit_tests.jl
+++ b/test/unit_tests/unit_tests.jl
@@ -1,5 +1,6 @@
 # Run all unit tests
 using LinearLeastSquares
+using Base.Test
 
 include("helpers.jl")
 include("add_subtract.jl")
@@ -8,3 +9,9 @@ include("stack.jl")
 include("diag.jl")
 include("repmat.jl")
 include("index.jl")
+include("reshape.jl")
+include("transpose.jl")
+include("sum_squares.jl")
+include("compound.jl")
+include("constraints.jl")
+include("problems.jl")
diff --git a/tests/test.jl b/tests/test.jl
deleted file mode 100644
index 5ed8b11..0000000
--- a/tests/test.jl
+++ /dev/null
@@ -1,46 +0,0 @@
-using LinearLeastSquares
-
-TOLERANCE = 1e-4;
-
-x = Variable(3);
-y = sum_squares(x);
-optval = minimize!(y);
-@assert abs(optval - 0) < TOLERANCE
-
-x = Variable(3);
-y = sum_squares(x);
-optval = minimize!(y, x == 3);
-@assert abs(optval - 27) < TOLERANCE
-
-x = Variable(2);
-A = [1 0; 1 0];
-y = sum_squares(x);
-optval = minimize!(y, A * x == [1; 2]);
-@assert optval == nothing
-
-A = randn(5, 5);
-x_real = randn(5, 5);
-x = Variable(5, 5);
-y = A * x;
-b = A * x_real;
-constraints = EqConstraint[];
-for i = 1:5
-  constraints += y[:, i] == b[:, i];
-end
-optval = minimize!(sum_squares(x), constraints);
-
-n = 200;
-# Specify the true value of the variable
-true_vect = [-1; 1];
-# Create data and labels
-X = randn(n, 2) * 2;
-b = randn(n, 1) * 0.1;
-y = X * true_vect + b;
-a = Variable(2);
-optval = minimize!(sum_squares(X * a - y));
-
-
-# x = Variable(5);
-# p = minimize!(sum_squares(sum(x)), x == 3);
-# solve!(p);
-# @assert abs(p.optval - 225) < TOLERANCE
diff --git a/tests/unit_tests/add_subtract.jl b/tests/unit_tests/add_subtract.jl
deleted file mode 100644
index a6cbaca..0000000
--- a/tests/unit_tests/add_subtract.jl
+++ /dev/null
@@ -1,9 +0,0 @@
-a = Variable();
-b = eye(3);
-c = a + b;
-set_value!(a, 1);
-val1 = eval_internals(c, (a,));
-val2 = c.evaluate();
-@assert all(val1 .== val2)
-
-info("All add/subtract tests passed")
diff --git a/tests/unit_tests/diag.jl b/tests/unit_tests/diag.jl
deleted file mode 100644
index 920bd09..0000000
--- a/tests/unit_tests/diag.jl
+++ /dev/null
@@ -1,33 +0,0 @@
-# diagm test
-x = Variable(4);
-d = diagm(x);
-set_value!(x, [1; 2; 3; 4]);
-val1 = eval_internals(d, (x,));
-val2 = d.evaluate();
-@assert all(val1 .== val2);
-
-# diag test 1
-x = Variable(3, 4);
-d = diag(x, 0);
-set_value!(x, [1 2 3 4; 5 6 7 8; 9 10 11 12]);
-val1 = eval_internals(d, (x,))
-val2 = d.evaluate();
-@assert all(val1 .== val2);
-
-# diag test 2
-x = Variable(3, 4);
-d = diag(x, 2);
-set_value!(x, [1 2 3 4; 5 6 7 8; 9 10 11 12]);
-val1 = eval_internals(d, (x,))
-val2 = d.evaluate();
-@assert all(val1 .== val2);
-
-# diag test 3
-x = Variable(3, 4);
-d = diag(x, -2);
-set_value!(x, [1 2 3 4; 5 6 7 8; 9 10 11 12]);
-val1 = eval_internals(d, (x,))
-val2 = d.evaluate();
-@assert all(val1 .== val2);
-
-info("All diag tests passed.")
diff --git a/tests/unit_tests/helpers.jl b/tests/unit_tests/helpers.jl
deleted file mode 100644
index c9a7564..0000000
--- a/tests/unit_tests/helpers.jl
+++ /dev/null
@@ -1,20 +0,0 @@
-export eval_internals, set_value!
-
-function eval_internals(x::AffineExpr, vars::Tuple)
-  (rows, cols) = x.size
-  value = zeros(rows, cols)
-  for var in vars
-    coeff = x.vars_to_coeffs_map[object_id(var)]
-    value += reshape(coeff.value * vec(var.value), rows, cols)
-  end
-  value += reshape(x.constant.value, rows, cols)
-  return value
-end
-
-function set_value!(x::AffineExpr, value::Number)
-  x.value = convert(Float64, value)
-end
-
-function set_value!(x::AffineExpr, value::AbstractArray)
-  x.value = convert(SparseMatrixCSC{Float64, Int64}, sparse(value))
-end
diff --git a/tests/unit_tests/index.jl b/tests/unit_tests/index.jl
deleted file mode 100644
index f0f8564..0000000
--- a/tests/unit_tests/index.jl
+++ /dev/null
@@ -1,13 +0,0 @@
-# indexing tests
-x = Variable(4, 6);
-set_value!(x, [1 2 3 4 5 6; 7 8 9 10 11 12; 13 14 15 16 17 18; 19 20 21 22 23 24]);
-y = x[1:3, 5:end];
-z = x[:];
-val1 = eval_internals(y, (x,));
-val2 = eval_internals(z, (x,));
-val3 = y.evaluate();
-val4 = z.evaluate();
-@assert all(val1 .== val3)
-@assert all(val2 .== val4)
-
-info("All index tests passed.")
diff --git a/tests/unit_tests/multiply_divide.jl b/tests/unit_tests/multiply_divide.jl
deleted file mode 100644
index df5f5e9..0000000
--- a/tests/unit_tests/multiply_divide.jl
+++ /dev/null
@@ -1,9 +0,0 @@
-a = Variable();
-b = eye(3);
-c = a * b;
-set_value!(a, 1);
-val1 = eval_internals(c, (a,));
-val2 = c.evaluate();
-@assert all(val1 .== val2)
-
-info("All multiply/divide tests passed")
diff --git a/tests/unit_tests/repmat.jl b/tests/unit_tests/repmat.jl
deleted file mode 100644
index 1363ac1..0000000
--- a/tests/unit_tests/repmat.jl
+++ /dev/null
@@ -1,32 +0,0 @@
-# repmat test 1
-x = Variable(3, 2);
-r = repmat(x, 4, 5);
-set_value!(x, [5 8; 6 9; 7 0]);
-val1 = eval_internals(r, (x,));
-val2 = r.evaluate();
-@assert all(val1 .== val2)
-
-x = Variable();
-r = repmat(x, 1, 1);
-set_value!(x, 3);
-val1 = eval_internals(r, (x,));
-val2 = r.evaluate();
-@assert all(val1 .== val2)
-
-x = Variable(3, 3);
-y = x + [1 2 3; 4 5 6; 7 8 9];
-r = repmat(y, 1, 3);
-set_value!(x, eye(3));
-val1 = eval_internals(r, (x,));
-val2 = r.evaluate();
-@assert all(val1 .== val2)
-
-x = Variable(3, 3);
-y = x + [1 2 3; 4 5 6; 7 8 9];
-r = repmat(y, 4, 1);
-set_value!(x, eye(3));
-val1 = eval_internals(r, (x,));
-val2 = r.evaluate();
-@assert all(val1 .== val2)
-
-info("All repmat tests passed.")
diff --git a/tests/unit_tests/stack.jl b/tests/unit_tests/stack.jl
deleted file mode 100644
index 96922db..0000000
--- a/tests/unit_tests/stack.jl
+++ /dev/null
@@ -1,29 +0,0 @@
-# hcat test
-x = Variable(3);
-y = Variable(3, 2);
-z = Variable(3, 3);
-h = [x y z];
-set_value!(x, [2; 3; 4]);
-set_value!(y, [5 8; 6 9; 7 0]);
-set_value!(z, eye(3));
-val1 = eval_internals(h, (x, y, z));
-val2 = h.evaluate();
-@assert all(val1 .== val2);
-
-# vcat test
-x = Variable(1, 3);
-y = Variable(2, 3);
-z = Variable(3, 3);
-v = [x; y; z];
-set_value!(x, [2 3 4]);
-set_value!(y, [5 6 7; 8 9 0]);
-set_value!(z, eye(3));
-val1 = eval_internals(v, (x, y, z));
-val2 = v.evaluate();
-@assert all(val1 .== val2);
-
-# hvcat test
-# TODO: add when hvcat is implemented
-
-
-info("All stack tests passed.")