Improved status messages and progress for the timeseries and CCA mode…

…ls. Closes #54.

packages/slycat/web/server/worker/model/__init__.py

@@ -228,7 +228,6 @@ def post_model_send_timeseries_rows(self, arguments):
       name = arguments["name"]
       rows = arguments["rows"]
       self.send_timeseries_artifact_binary_rows(name, rows)
-      self.set_message("Timeseries %s stored %s rows." % (name, self.artifacts[name].row_count))
     except KeyError as e:
       raise cherrypy.HTTPError("400 Missing key: %s" % e.message)
 
@@ -365,7 +364,6 @@ def __init__(self, column_names, column_types):
     self.named_pipe = None
     self.stream = None
     self.thread = None
-    self.row_count = 0
 
   def get_stream(self):
     def load_data(database, array, named_pipe, column_types):
@@ -393,7 +391,6 @@ def store_rows(self, rows):
           stream.write("\0")
         elif type == "double":
           stream.write(struct.pack("<d", value))
-      self.row_count += 1
 
   def store_columns(self, row_count, columns):
     stream = self.get_stream()
@@ -407,7 +404,6 @@ def store_columns(self, row_count, columns):
           stream.write("\0")
         elif self.column_types[column] == "double":
           stream.write(struct.pack("<d", value))
-      self.row_count += 1
 
   def finish(self):
     if self.stream is not None:
@@ -436,7 +432,6 @@ def __init__(self, column_names, column_types):
     self.named_pipe = None
     self.stream = None
     self.thread = None
-    self.row_count = 0
 
   def get_stream(self):
     def load_data(database, array, named_pipe, column_types):
@@ -465,7 +460,6 @@ def store_rows(self, ids, times, rows):
           stream.write("\0")
         elif type == "double":
           stream.write(struct.pack("<d", value))
-      self.row_count += 1
 
   def finish(self):
     if self.stream is not None:
@@ -501,7 +495,6 @@ def __init__(self, attributes, dimensions):
     self.named_pipe = None
     self.stream = None
     self.thread = None
-    self.row_count = 0
 
   def get_stream(self):
     def load_data(database, array, named_pipe, attribute_types):

packages/slycat/web/server/worker/model/cca3.py

@@ -16,6 +16,9 @@ def __init__(self, security, pid, mid, name, marking, description):
     slycat.web.server.worker.model.prototype.__init__(self, security, "CCA model", pid, mid, "cca3", name, marking, description, incremental=True)
 
   def compute_model(self):
+    self.set_progress(0.0)
+    self.set_message("Transforming data.")
+
     # Get required inputs ...
     data_table = self.load_table_artifact("data-table")
     input_columns = self.load_json_artifact("input-columns")
@@ -45,6 +48,7 @@ def compute_model(self):
 #    cherrypy.log.error("%s" % X)
 #    cherrypy.log.error("%s" % Y)
 #    cherrypy.log.error("%s" % scale_inputs)
+    self.set_message("Computing CCA.")
     x, y, x_loadings, y_loadings, r, wilks = cca(X, Y, scale_inputs=scale_inputs)
 #    cherrypy.log.error("%s" % x)
 #    cherrypy.log.error("%s" % y)
@@ -53,6 +57,7 @@ def compute_model(self):
 #    cherrypy.log.error("%s" % r)
 #    cherrypy.log.error("%s" % wilks)
 
+    self.set_message("Storing results.")
     component_count = x.shape[1]
     sample_count = x.shape[0]
 
@@ -79,3 +84,4 @@ def compute_model(self):
       self.send_array_artifact_data("cca-statistics", [r[component], wilks[component]])
     self.finish_array_artifact("cca-statistics", input=False)
 
+    self.set_progress(1.0)

packages/slycat/web/server/worker/model/timeseries.py

@@ -35,9 +35,13 @@ def compute_model(self):
     self.store_json_file_artifact("clusters", sorted([name for name, artifact in cluster_artifacts]), input=False)
 
     # For each cluster (timeseries set) ...
-    for name, artifact in cluster_artifacts:
-      # Rebin the timeseries within this cluster so they share common start / stop times and samples ...
+    for index, (name, artifact) in enumerate(cluster_artifacts):
+      progress_begin = float(index) / float(len(cluster_artifacts))
+      progress_end = float(index + 1) / float(len(cluster_artifacts))
+      self.set_progress(progress_begin)
+
       self.set_message("Rebinning data for %s" % name)
+      # Rebin the timeseries within this cluster so they share common start / stop times and samples ...
       time_min = self.scidb.query_value("aql", "select min(time) from %s" % artifact["columns"]).getDouble()
       time_max = self.scidb.query_value("aql", "select max(time) from %s" % artifact["columns"]).getDouble()
 
@@ -73,6 +77,7 @@ def compute_model(self):
       observation_count = len(cluster)
       distance_matrix = numpy.zeros(shape=(observation_count, observation_count))
       for i in range(0, observation_count):
+        self.set_progress(self.mix(progress_begin, progress_end, float(i) / float(observation_count)))
         self.set_message("Computing distance matrix for %s, %s of %s" % (name, i+1, observation_count))
         for j in range(i + 1, observation_count):
           distance = math.sqrt(math.fsum([math.pow(b - a, 2.0) for a, b in zip(cluster[i]["values"], cluster[j]["values"])]))