Improving csv -> dataframe module

[predictops.git] / main.py

diff --git a/main.py b/main.py
index 6810673765688db13b0a2b8ec6e3cef5ff53bb1f..a42ce0f11cb9b42feae17fa17e1abcdf513b8570 100644 (file)
--- a/main.py
+++ b/main.py
@@ -1,24 +1,25 @@
-from lib.source import MeteoFrance
+from predictops.source.ephemeris import Ephemeris
+from predictops.source.meteofrance import MeteoFrance
+from predictops.learn.preprocessing import Preprocessing
+from predictops.target.toarea import ToArea

 
 

-from configparser import ConfigParser
-from logging.config import fileConfig
+from datetime import datetime, timedelta

 from logging import getLogger
 from logging import getLogger

+from logging.config import fileConfig

 from pathlib import Path
 from shutil import rmtree
 from pathlib import Path
 from shutil import rmtree

-from subprocess import Popen, PIPE

 
 

+import geopandas as gpd

 
 fileConfig((Path.cwd() / 'config') / 'logging.cfg')
 logger = getLogger()
 
 
 class Engine:
 
 fileConfig((Path.cwd() / 'config') / 'logging.cfg')
 logger = getLogger()
 
 
 class Engine:

-    def __init__(self, clean = False):
-        logger.info("Predictops engine launched")
-        if clean:
-            self.clean()
-            print("Ne pas oublier d'exporter la BDD dans pgModeler")
-            print("Ni de copier l'archive dans la data")
+    def __init__(self, start = None, end = None, time_step = None):
+        self._X = {}
+        self._Y = {}
+

 
     def clean(self):
         # Cleaning the data directory
 
     def clean(self):
         # Cleaning the data directory


@@ -29,29 +30,60 @@ class Engine:
         p = Path(Path.cwd() / 'data')
         p.mkdir()
 
         p = Path(Path.cwd() / 'data')
         p.mkdir()
 

-        # Cleaning the postgresql database
-        config = ConfigParser()
-        config.read((Path.cwd() / 'config') / 'main.cfg')

 
 

-        host   = config['postgresql']['host']
-        user   = config['postgresql']['user']
-        port   = config['postgresql']['port']
-        dbname = config['postgresql']['dbname']
+    def add_feature(self, name, **kw):
+
+        if name == 'meteofrance':
+            meteofeature = MeteoFrance(**kw)
+            meteofeature.update()
+            dated_features = meteofeature.dated_features
+            for date in dated_features:
+                self._X.setdefault(date,{}).update(dated_features[date])
+        elif name == 'ephemeris':
+            ephemerides = Ephemeris(**kw)
+            dated_features = ephemerides.dated_features
+            for date in dated_features:
+                self._X.setdefault(date,{}).update(dated_features[date])
+
+
+    @property
+    def X(self):
+        return self._X
+
+    @X.setter
+    def X(self, x):
+        self._X = x
+
+
+start = datetime.strptime('01/01/2010 01:00:00', '%m/%d/%Y %H:%M:%S')
+end = datetime.strptime('12/31/2010 23:00:00', '%m/%d/%Y %H:%M:%S')
+
+engine = Engine()
+engine.add_feature(name = 'meteofrance',
+                   start = start, end = end,
+                   latitude = 47.25, longitude = 6.0333, nb_stations = 3,
+                   features = ['temperature', 'pressure'])
+
+
+engine.add_feature(name = 'ephemeris',
+                   start = start, end = end,
+                   features = ['hour', 'dayInWeek', 'dayInMonth', 'dayInYear',
+                               'weekInYear', 'month', 'year'])
+

 
 

-        logger.info("PostgreSQL database deletion")
-        command = ['dropdb', '-h', host, '-U', user, '-p', port, dbname]
-        process = Popen(command, stdout=PIPE, stderr=PIPE)
-        process.communicate()
+process = Preprocessing(dict_features = engine.X,
+                   start = start, end = end, timestep = timedelta(hours=6))

 
 

-        logger.info("PostgreSQL database creation")
-        command = ['createdb', '-h', host, '-U', user, '-p', port, dbname]
-        process = Popen(command, stdout=PIPE, stderr=PIPE)
-        process.communicate()

 
 

-    def add_meteofrance(self):
-        self.meteofrance = MeteoFrance()
+df = process.dataframe.head(n=20)
+#print(engine.X[datetime.strptime('06/30/2010 21:00:00', '%m/%d/%Y %H:%M:%S')])
+print(df)
+exit()

 
 

+depts = gpd.read_file( Path.cwd() / 'data' / 'targets' / 'departments' / "departements-20180101.shp")
+Doubs = depts.loc[depts['nom'] == 'Doubs'].iloc[0]

 
 

+ToArea(area=Doubs.geometry,
+       start = start, end = end,
+       csv_file = Path.cwd() / 'data' / 'targets' / 'sdis25' / 'interventions.csv')

 
 

-engine = Engine(clean = False)
-engine.add_meteofrance()
\ No newline at end of file