Use UTC datetimes everywhere

Now datetime math is done in UTC for simplicity. Furthermore, created_on_str is
always UTC. Finally, a new migration for created_on and modified_on fields was
created.

Author: rubik

history/admin.py

@@ -26,9 +26,9 @@ def algo(self,obj):
         if trs.count:
             tr = trs[0]
             if tr.clf:
-                html +=  "<a href="https://github.com/admin/history/classifiertest/{}">{}</a>".format(tr.clf.pk,tr.clf) 
+                html +=  "<a href="https://github.com/admin/history/classifiertest/{}">{}</a>".format(tr.clf.pk,tr.clf)
             if tr.made_by:
-                html +=  "<a href="https://github.com/admin/history/predictiontest/{}">{}</a>".format(tr.made_by.pk,tr.made_by) 
+                html +=  "<a href="https://github.com/admin/history/predictiontest/{}">{}</a>".format(tr.made_by.pk,tr.made_by)
         return html
 
     algo.allow_tags = True
@@ -82,6 +82,3 @@ def view_link(obj):
     list_display = ['pk', 'type','symbol', 'name', 'created_on', 'percent_correct', 'score', 'prediction_size', view_link]
 
 admin.site.register(ClassifierTest, ClassifierTestAdmin)
-
-
-

history/management/commands/alert_fail_cases.py

@@ -1,7 +1,8 @@
-from django.core.management.base import BaseCommand
 import datetime
-from history.models import PredictionTest, TradeRecommendation, get_time
+from django.utils import timezone
 from django.conf import settings
+from django.core.management.base import BaseCommand
+from history.models import PredictionTest, TradeRecommendation
 
 class Command(BaseCommand):
 
@@ -35,9 +36,10 @@ def handle(self, *args, **options):
         print(last_pt.created_on)
         print(last_trade.created_on)
 
-        # 7 hours thing is a hack for MST vs UTC timezone issues
-        is_trader_running = last_trade.created_on > (get_time() - datetime.timedelta(hours=int(7)) - datetime.timedelta(minutes=int(15)))
-        is_trainer_running = last_pt.created_on > (get_time() - datetime.timedelta(hours=int(7)) - datetime.timedelta(minutes=int(15)))
+        # Since USE_TZ=True, created_on is already UTC, like timezone.now()
+        fifteen_ago = timezone.now() - datetime.timedelta(minutes=15)
+        is_trader_running = last_trade.created_on > fifteen_ago
+        is_trainer_running = last_pt.created_on > fifteen_ago
 
         if not is_trader_running:
             self.alert_email("not is_trader_running")

history/management/commands/trade.py

@@ -1,6 +1,7 @@
 from django.core.management.base import BaseCommand
 from history.predict import predict_v2
-from history.models import Price, PredictionTest, Trade, TradeRecommendation, Balance, get_time, ClassifierTest
+from history.models import (Price, PredictionTest, Trade, TradeRecommendation,
+                            Balance, ClassifierTest)
 from history.tools import get_utc_unixtime, print_and_log
 import datetime
 import time
@@ -186,7 +187,7 @@ def run_predictor(self, nn_index):
                                  confidence=confidence,
                                  recommendation=recommend,
                                  net_amount=-1 if recommend == 'SELL' else (1 if recommend == 'BUY' else 0),
-                                 created_on_str=str(get_time().strftime('%Y-%m-%d %H:%M')))
+                                 created_on_str=timezone.now().strftime('%Y-%m-%d %H:%M'))
         tr.save()
         self.trs[nn_index] = tr
         return recommend

history/migrations/0004_auto_20160409_1213.py

@@ -0,0 +1,94 @@
+# -*- coding: utf-8 -*-
+from __future__ import unicode_literals
+
+from django.db import models, migrations
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('history', '0003_auto_20160330_1920'),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name='balance',
+            name='created_on',
+            field=models.DateTimeField(auto_now_add=True, db_index=True),
+        ),
+        migrations.AlterField(
+            model_name='balance',
+            name='modified_on',
+            field=models.DateTimeField(auto_now=True),
+        ),
+        migrations.AlterField(
+            model_name='classifiertest',
+            name='created_on',
+            field=models.DateTimeField(auto_now_add=True),
+        ),
+        migrations.AlterField(
+            model_name='classifiertest',
+            name='modified_on',
+            field=models.DateTimeField(auto_now=True),
+        ),
+        migrations.AlterField(
+            model_name='deposit',
+            name='created_on',
+            field=models.DateTimeField(auto_now_add=True, db_index=True),
+        ),
+        migrations.AlterField(
+            model_name='deposit',
+            name='modified_on',
+            field=models.DateTimeField(auto_now=True),
+        ),
+        migrations.AlterField(
+            model_name='performancecomp',
+            name='created_on',
+            field=models.DateTimeField(auto_now_add=True, db_index=True),
+        ),
+        migrations.AlterField(
+            model_name='performancecomp',
+            name='modified_on',
+            field=models.DateTimeField(auto_now=True),
+        ),
+        migrations.AlterField(
+            model_name='predictiontest',
+            name='created_on',
+            field=models.DateTimeField(auto_now_add=True),
+        ),
+        migrations.AlterField(
+            model_name='predictiontest',
+            name='modified_on',
+            field=models.DateTimeField(auto_now=True),
+        ),
+        migrations.AlterField(
+            model_name='price',
+            name='created_on',
+            field=models.DateTimeField(auto_now_add=True, db_index=True),
+        ),
+        migrations.AlterField(
+            model_name='price',
+            name='modified_on',
+            field=models.DateTimeField(auto_now=True),
+        ),
+        migrations.AlterField(
+            model_name='trade',
+            name='created_on',
+            field=models.DateTimeField(auto_now_add=True, db_index=True),
+        ),
+        migrations.AlterField(
+            model_name='trade',
+            name='modified_on',
+            field=models.DateTimeField(auto_now=True),
+        ),
+        migrations.AlterField(
+            model_name='traderecommendation',
+            name='created_on',
+            field=models.DateTimeField(auto_now_add=True, db_index=True),
+        ),
+        migrations.AlterField(
+            model_name='traderecommendation',
+            name='modified_on',
+            field=models.DateTimeField(auto_now=True),
+        ),
+    ]

history/models.py

@@ -27,15 +27,16 @@
 from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis
 
 import warnings
-warnings.filterwarnings("ignore", category=DeprecationWarning) 
+warnings.filterwarnings("ignore", category=DeprecationWarning)
 
 
 def get_time():
-    return localtime(timezone.now())    
+    return timezone.localtime(timezone.now())
+
 
 class TimeStampedModel(models.Model):
-    created_on = models.DateTimeField(null=False, default=get_time, db_index=True)
-    modified_on = models.DateTimeField(null=False, default=get_time)
+    created_on = models.DateTimeField(null=False, auto_now_add=True, db_index=True)
+    modified_on = models.DateTimeField(null=False, auto_now=True)
 
     def get_readonly_fields(self, request, obj=None):
         return [f.name for f in self._meta.get_fields()]
@@ -46,10 +47,6 @@ def has_add_permission(self, request, obj=None):
     def has_delete_permission(self, request, obj=None):
         return False
 
-    def save(self, *args, **kwargs):
-        self.modified_on = get_time()
-        return super(TimeStampedModel, self).save(*args, **kwargs)
-
     class Meta:
         abstract = True
 
@@ -58,8 +55,8 @@ def url_to_edit_object(self):
         return '<a href="{0}">Edit {1}</a>'.format(url, escape(str(self)))
 
 class AbstractedTesterClass(models.Model):
-    created_on = models.DateTimeField(null=False, default=get_time)
-    modified_on = models.DateTimeField(null=False, default=get_time)
+    created_on = models.DateTimeField(null=False, auto_now_add=True)
+    modified_on = models.DateTimeField(null=False, auto_now=True)
 
     def get_readonly_fields(self, request, obj=None):
         return [f.name for f in self._meta.get_fields()]
@@ -70,17 +67,13 @@ def has_add_permission(self, request, obj=None):
     def has_delete_permission(self, request, obj=None):
         return False
 
-    def save(self, *args, **kwargs):
-        self.modified_on = get_time()
-        return super(AbstractedTesterClass, self).save(*args, **kwargs)
-
     class Meta:
         abstract = True
 
     def url_to_edit_object(self):
         url = reverse('admin:{0}_{1}_change'.format(self._meta.app_label, self._meta.model_name), args=[self.id])
         return '<a href="{0}">Edit {1}</a>'.format(url, escape(str(self)))
-    
+
     def confidence(self):
         related = self.related_mocks()
         related.exclude(percent_correct__isnull=True)
@@ -221,7 +214,7 @@ class PerformanceComp(TimeStampedModel):
     price_timerange_end = models.DateTimeField(null=True, default=None,db_index=True)
     tr_timerange_start = models.DateTimeField(null=True, default=None)
     tr_timerange_end = models.DateTimeField(null=True, default=None)
-    
+
 
 class TradeRecommendation(TimeStampedModel):
     symbol = models.CharField(max_length=30)
@@ -313,7 +306,7 @@ def get_classifier(self,train=True,test=True):
                     price_datasets[1].append(do_buy)
                 except Exception as e:
                     pass
-                    
+
             data = price_datasets
             if self.timedelta_back_in_granularity_increments == 0:
                 train_data = data
@@ -338,7 +331,7 @@ def get_classifier(self,train=True,test=True):
 
             clf.fit(self.X_train, self.y_train)
             score = clf.score(self.X_test, self.y_test)
-            
+
             # Plot the decision boundary. For that, we will assign a color to each
             # point in the mesh [self.x_min, m_max]x[self.y_min, self.y_max].
 
@@ -347,7 +340,7 @@ def get_classifier(self,train=True,test=True):
                 self.ravel_args.append(self.xz[i].ravel())
 
             self._input = np.column_stack(self.ravel_args)
-            
+
             if hasattr(clf, "decision_function"):
                 self.Z = clf.decision_function(self._input)
             else:
@@ -370,7 +363,7 @@ def get_classifier(self,train=True,test=True):
                 all_output = all_output + str(('stats_debug',stats))
                 self.percent_correct = int(pct_correct*100)
                 self.prediction_size = len(test_data[0])
-            
+
             all_output = all_output + str((self.name,round(score*100)))
             self.score = score*100
             end_time = int(time.time())
@@ -400,7 +393,7 @@ def graph(self,filename):
         figure = plt.figure(figsize=(27, 9))
         figure.max_num_figures = 5
         matplotlib.figure.max_num_figures = 5
-        i = 0 
+        i = 0
         cm = plt.cm.RdBu
         cm_bright = ListedColormap(['#00FF00','#FF0000', '#0000FF'])
         ax = plt.subplot(1, 1 , i)
@@ -530,9 +523,9 @@ def get_nn(self,train=True):
         except ImportError:
             FNN = buildNetwork(DS.indim, self.hiddenneurons, DS.outdim, bias=self.bias, recurrent=self.recurrent)
         FNN.randomize()
-        
+
         TRAINER = BackpropTrainer(FNN, dataset=DS, learningrate = self.learningrate, \
-            momentum=self.momentum, verbose=False, weightdecay=self.weightdecay) 
+            momentum=self.momentum, verbose=False, weightdecay=self.weightdecay)
 
         if train:
             for i in range(self.epochs):
@@ -546,12 +539,12 @@ def recommend_trade(self,nn_price,last_sample, fee_amount = get_fee_amount()):
         fee_amount = fee_amount * settings.FEE_MANAGEMENT_STRATEGY # see desc in settings.py
         anticipated_percent_increase = (nn_price - last_sample) / last_sample
         if abs(anticipated_percent_increase) < fee_amount:
-            should_trade = 'HOLD' 
+            should_trade = 'HOLD'
         elif anticipated_percent_increase > fee_amount:
-            should_trade = 'BUY' 
+            should_trade = 'BUY'
         elif anticipated_percent_increase < fee_amount:
             should_trade = 'SELL'
-        return should_trade 
+        return should_trade
 
     def predict(self,sample):
         last_sample = sample[-1]