updated distplot example; added conditional to block server script ex…

…it when exception occurs

examples/for_seaborn/distplot.cpp

@@ -12,46 +12,57 @@
 
 int main()
 {
-  Eigen::Matrix<double, 1000, 1> mat;
-  mat.setRandom();
-
   Cppyplot::cppyplot pyp;
 
-  pyp.raw(R"pyp(
-    fig = plt.figure(figsize=(6,5))
-    ax = sns.distplot(mat, kde=False,
-                      hist_kws={"histtype": "bar", "rwidth": 3,
-                                "alpha": 0.6, "color": "b"})
-    sns.distplot(mat, kde=False,
-                 hist_kws={"histtype": "step", "rwidth": 2,
-                           "alpha": 0.9, "color": "orange"})
-    ax.set_xlabel("Data bins", fontsize=12)
-    ax.set_ylabel("Frequency", fontsize=12)
-    ax.set_title("Histogram of C++ Eigen container", fontsize=14)
-    plt.show()
-  )pyp");
-  pyp.data_args(_p(mat));
+  Eigen::Matrix<double, 5000, 2> mat;
 
   // bivariate plot from https://seaborn.pydata.org/examples/layered_bivariate_plot.html
 
-  std::vector<std::vector<double>> rand_mat (1000, std::vector<double>(2));
+  std::vector<std::vector<double>> rand_mat (5000, std::vector<double>(2));
   std::random_device seed;
   std::mt19937 gen(seed());
-  std::normal_distribution<double> norm(0.0, 10.0);
+  std::normal_distribution<double> norm(0.0, 5.0);
 
-  // fill matrix
+  // fill 2D vector
   for (auto& row: rand_mat)
   {
     for (auto& elem: row)
     { elem = norm(gen); }
   }
 
+  // fill eigen matrix
+  for (auto& row: mat.rowwise())
+  {
+    for (auto& elem : row)
+    { elem = norm(gen); }
+  }
+
+  // Using Eigen matrix
   pyp.raw(R"pyp(
   sns.set_theme(style="dark")
   f, ax = plt.subplots(figsize=(6,6))
+  sns.scatterplot(x=mat[:,0], y=mat[:,1], s=5, color="0.15")
+  sns.histplot(x=mat[:,0], y=mat[:,1], bins=50, pthresh=0.1, cmap="mako")
+  sns.kdeplot(x=mat[:,0], y=mat[:,1], levels=5, color="w", linewidths=1)
+  plt.grid(True)
+  plt.xlabel("X", fontsize=12)
+  plt.ylabel("Y", fontsize=12)
+  plt.title("Numpy type slicing on C++ Eigen matrix", fontsize=14)
+  plt.show(block=False)
+    )pyp");
+
+  pyp.data_args(_p(mat));
+
+  // Using 2D std-vector
+  pyp.raw(R"pyp(
+  f, ax = plt.subplots(figsize=(6,6))
   sns.scatterplot(x=rand_mat[:,0], y=rand_mat[:,1], s=5, color="0.15")
   sns.histplot(x=rand_mat[:,0], y=rand_mat[:,1], bins=50, pthresh=0.1, cmap="mako")
   sns.kdeplot(x=rand_mat[:,0], y=rand_mat[:,1], levels=5, color="w", linewidths=1)
+  plt.grid(True)
+  plt.xlabel("X", fontsize=12)
+  plt.ylabel("Y", fontsize=12)
+  plt.title("Numpy type slicing on C++ 2D vector", fontsize=14)
   plt.show()
     )pyp");
 

include/cppyplot_server.py

@@ -26,6 +26,10 @@
 # import seaborn as sns
 # lib_sym['sns'] = sns
 
+## Import pandas and register pandas object in the symbol table
+# import pandas as pd
+# lib_sym['pd'] = pd
+
 # Import bokeh and register bokeh objects in the symbol table
 # from bokeh.layouts import column, row
 # lib_sym['column'] = column
@@ -69,11 +73,16 @@ def parse_shape(shape_str:str)->tuple:
         data_payload  = socket.recv()
         data_payload  = np.array((struct.unpack("="+(data_type*data_len), data_payload)))
         plot_data[data_info[1]] = np.reshape(data_payload, data_shape)
-
     elif(zmq_message[0:8] == b"finalize"):
         sym_table = {**lib_sym, **plot_data}
         aeval = Interpreter(usersyms=sym_table)
         aeval.eval(plot_cmd)
+
+        # Some error happened pause execution by creating sample matplotlib windows
+        if aeval.error_msg != None:
+            plt.figure(figsize=(6,5))
+            plt.title("Exception from ASTEVAL, check stdout", fontsize=14)
+            plt.show()
         plot_cmd  = None
         plot_data = {}
     elif(zmq_message == b"exit"):