pwfts.visualize_distributions

pyFTS/models/multivariate/flrg.py

@@ -10,14 +10,21 @@ class FLRG(flg.FLRG):
 
     def __init__(self, **kwargs):
         super(FLRG,self).__init__(0,**kwargs)
+        self.order = kwargs.get('order', 1)
         self.LHS = kwargs.get('lhs', {})
         self.RHS = set()
 
-    def set_lhs(self, var, set):
+    def set_lhs(self, var, fset):
-        self.LHS[var] = set
+        if self.order == 1:
+            self.LHS[var] = fset
+        else:
+            if var not in self.LHS:
+                self.LHS[var] = []
+            self.LHS[var].append(fset)
 
-    def append_rhs(self, set, **kwargs):
+
-        self.RHS.add(set)
+    def append_rhs(self, fset, **kwargs):
+        self.RHS.add(fset)
 
     def get_membership(self, data, variables):
         mvs = []

pyFTS/models/multivariate/mvfts.py

@@ -91,6 +91,8 @@ class MVFTS(fts.FTS):
 
         ndata = self.apply_transformations(data)
 
+        self.order = kwargs.get('order',1)
+
         flrs = self.generate_flrs(ndata)
         self.generate_flrg(flrs)
 

pyFTS/models/pwfts.py

@@ -517,3 +517,39 @@ class ProbabilisticWeightedFTS(ifts.IntervalFTS):
             p = round(self.flrgs[r].frequency_count / self.global_frequency_count, 3)
             tmp = tmp + "(" + str(p) + ") " + str(self.flrgs[r]) + "\n"
         return tmp
+
+
+def visualize_distributions(model, **kwargs):
+    import matplotlib.pyplot as plt
+    from matplotlib import gridspec
+    import seaborn as sns
+
+    ordered_sets = model.partitioner.ordered_sets
+    ftpg_keys = sorted(model.flrgs.keys(), key=lambda x: model.flrgs[x].get_midpoint(model.sets))
+
+    lhs_probs = [model.flrg_lhs_unconditional_probability(model.flrgs[k])
+                 for k in ftpg_keys]
+
+    mat = np.zeros((len(ftpg_keys), len(ordered_sets)))
+    for row, w in enumerate(ftpg_keys):
+        for col, k in enumerate(ordered_sets):
+            if k in model.flrgs[w].RHS:
+                mat[row, col] = model.flrgs[w].rhs_unconditional_probability(k)
+
+    size = kwargs.get('size', (5,10))
+
+    fig = plt.figure(figsize=size)
+
+    gs = gridspec.GridSpec(1, 2, width_ratios=[1, 4])
+    ax1 = plt.subplot(gs[0])
+    sns.barplot(x='y', y='x', color='darkblue', data={'x': ftpg_keys, 'y': lhs_probs}, ax=ax1)
+    ax1.set_ylabel("LHS Probabilities")
+
+    ind_sets = range(len(ordered_sets))
+    ax = plt.subplot(gs[1])
+    sns.heatmap(mat, cmap='Blues', ax=ax, yticklabels=False)
+    ax.set_title("RHS probabilities")
+    ax.set_xticks(ind_sets)
+    ax.set_xticklabels(ordered_sets)
+    ax.grid(True)
+    ax.xaxis.set_tick_params(rotation=90)

pyFTS/tests/multivariate.py

@@ -6,7 +6,7 @@ from pyFTS.common import Transformations
 
 from pyFTS.data import SONDA
 df = SONDA.get_dataframe()
-train = df.iloc[0:1578241] #three years
+train = df.iloc[0:578241] #three years
 #test = df.iloc[1572480:2096640] #ears
 del df
 
@@ -26,36 +26,40 @@ model = mvfts.MVFTS("")
 
 fig, axes = plt.subplots(nrows=5, ncols=1,figsize=[15,10])
 
+
 sp = {'seasonality': DateTime.day_of_year , 'names': ['Jan','Feb','Mar','Apr','May','Jun','Jul', 'Aug','Sep','Oct','Nov','Dec']}
 
 vmonth = variable.Variable("Month", data_label="datahora", partitioner=seasonal.TimeGridPartitioner, npart=12,
                            data=train, partitioner_specific=sp)
-model.append_variable(vmonth)
+vmonth.partitioner.plot(axes[0])
 
 sp = {'seasonality': DateTime.minute_of_day}
 
 vhour = variable.Variable("Hour", data_label="datahora", partitioner=seasonal.TimeGridPartitioner, npart=24,
                           data=train, partitioner_specific=sp)
-model.append_variable(vhour)
 
-vhumid = variable.Variable("Humidity", data_label="humid", partitioner=Grid.GridPartitioner, npart=np, data=train,
+vhour.partitioner.plot(axes[1])
-                           transformation=tdiff)
-model.append_variable(vhumid)
 
-vpress = variable.Variable("AtmPress", data_label="press", partitioner=Grid.GridPartitioner, npart=np, data=train,
+vavg = variable.Variable("Radiance", data_label="glo_avg", partitioner=Grid.GridPartitioner, npart=30,
-                           transformation=tdiff)
+                         data=train)
-model.append_variable(vpress)
 
-vrain = variable.Variable("Rain", data_label="rain", partitioner=Grid.GridPartitioner, npart=20, data=train)#train)
+model1 = mvfts.MVFTS("")
-model.append_variable(vrain)
 
-model.target_variable = vrain
+model1.append_variable(vmonth)
+
+model1.append_variable(vhour)
+
+model1.append_variable(vavg)
+
+model1.target_variable = vavg
+
+#model1.fit(train, num_batches=60, save=True, batch_save=True, file_path='mvfts_sonda')
 
 
 #model.fit(train, num_batches=60, save=True, batch_save=True, file_path='mvfts_sonda')
 
-model.fit(train, num_batches=200, save=True, batch_save=True, file_path='mvfts_sonda', distributed=True,
+model1.fit(train, num_batches=200, save=True, batch_save=True, file_path='mvfts_sonda', distributed=True,
-          nodes=['192.168.1.22'], batch_save_interval=10)
+          nodes=['192.168.1.35'], batch_save_interval=10)
 
 
 #model = Util.load_obj('mvfts_sonda')