Update README.md

README.md

@@ -33,5 +33,5 @@ Visualization using Seaborn/Matplotlib
 1. [Tanishka Gaur](https://github.com/tanishkaa31/)
 2. [Vedant Chourasia](https://github.com/Vc8006/)
 3. [Shriya Dixit](https://github.com/shriyaD13/)
-4. [Pranjali Verma](https://github.com/pranjali1909/)
+4. [Pranjali Jadkar](https://github.com/pranjali1909/)
 5. [Shrish Kumar Singhal](https://github.com/shrishsinghal/)

main/Ml_model.py

@@ -0,0 +1,156 @@
+import pandas as pd
+import numpy as np
+from matplotlib import pyplot as plt
+from .models import Data , Data_Predicted ,Update
+from datetime import date
+import datetime
+import statsmodels.api as sm
+from pandas.tseries.offsets import DateOffset
+
+# df = pd.read_csv("./main/static/new_data.csv",parse_dates = ['Date'], index_col = ['Date'])
+update = Update()
+data1 = Data_Predicted()
+data = Data()
+q = Data.objects.all().values()
+df = pd.DataFrame(q)
+df['Date'] = pd.to_datetime(df['Date'])
+df = df.set_index(df['Date']) 
+
+m = Data_Predicted.objects.all().values()
+df_pred = pd.DataFrame(m)
+df_pred['Date'] = pd.to_datetime(df_pred['Date'])
+df_pred = df_pred.set_index(df_pred['Date']) 
+
+from sklearn.metrics import mean_squared_error
+import statsmodels.graphics.tsaplots as sgt
+
+from statsmodels.tsa.arima_model import ARIMA
+df = df_pred
+size =  int(len(df)*0.8)
+df_train = df.iloc[:size]
+df_test = df.iloc[size:]
+
+# u = Update.objects.get(name = "")
+u = Update.objects.all().values()
+
+date1 = [2555]
+date2 = [2556]
+def fun(df):
+    dfa = df['AQI']
+    df = df.dropna()
+    future_dates = [df_pred.index[-1] + DateOffset(days = x)for x in range(2)]
+    future_datest_df_new = pd.DataFrame(index = future_dates[1:],columns=df.columns)
+
+    my_order = (0, 0, 0)
+    my_seasonal_order = (1, 0, 1, 12)
+    smodel = sm.tsa.SARIMAX(dfa, order=my_order,seasonal_order=my_seasonal_order,trend= 't')
+    result_seasonal = smodel.fit()
+    # date1 = future_dates[0].date()
+    # date2 = future_dates[1].date()
+    predictions_seasonal = result_seasonal.predict(steps = 1)
+    # predictions_seasonal = result_seasonal.predict(start=date1[0],end=date2[0],typ="levels")
+    # predictions_seasonal = result_seasonal.forecast(steps = 1)[0]
+    # data1.AQI = 20
+    print('############################ f_date',future_dates[0].date())
+    print('############################ future dates',future_dates)
+    x_date_n = ""
+    x_date_n = future_dates[0].date().strftime("%Y-%m-%d")
+    print(x_date_n)
+    print('AQI       :*********************** ',predictions_seasonal)
+    data1.AQI = predictions_seasonal.iloc[-1]
+    # data.AQI = predictions_seasonal.iloc[-1]
+
+    # date1[0] = date1[0] + 1
+    # date2[0] = date2[0] + 1
+
+    # model = ARIMA(df_train['AQI'],order = (2,1,2))
+    # model_fit = model.fit()
+    # data1.AQI = model_fit.forecast(steps = 1)[0]
+    # future_datest_df_new['AQI'] = model_fit.forecast(steps = 1)[0]
+
+    smodel_so2 = sm.tsa.SARIMAX(df['SO2'], order=my_order,seasonal_order=my_seasonal_order,trend= 't')
+    result_seasonal_so2 = smodel_so2.fit()
+    predictions_seasonal_so2 = result_seasonal_so2.predict(steps = 1)
+    # predictions_seasonal_so2 = result_seasonal_so2.forecast(steps = 1)[0]
+    print('so2      : ',predictions_seasonal_so2.iloc[-1])
+    data1.SO2 = predictions_seasonal_so2.iloc[-1]
+    # data.SO2 = predictions_seasonal_so2.iloc[-1]
+    # data1.SO2 = 20
+
+    # model = ARIMA(df_train['SO2'],order = (2,1,2))
+    # model_fit = model.fit()
+    # # data1.SO2 = model_fit.forecast(steps = 1)[0]
+    # # future_datest_df_new['SO2'] = model_fit.forecast(steps = 1)[0]
+    # data1.SO2 = 20
+
+
+    smodel_no2 = sm.tsa.SARIMAX(df['NO2'], order=my_order,seasonal_order=my_seasonal_order,trend= 't')
+    result_seasonal_no2 = smodel_no2.fit()
+    predictions_seasonal_no2 = result_seasonal_no2.predict(steps = 1)
+    # predictions_seasonal_no2 = result_seasonal_no2.forecast(steps = 1)[0]
+    print('no2        : ',predictions_seasonal_no2.iloc[-1])
+    data1.NO2 = predictions_seasonal_no2.iloc[-1]
+    # data.NO2 = predictions_seasonal_no2.iloc[-1]
+    # data1.NO2 = 20
+
+    # model = ARIMA(df_train['NO2'],order = (2,1,2))
+    # model_fit = model.fit()
+    # data1.NO2 = model_fit.forecast(steps = 1)[0]
+    # future_datest_df_new['NO2'] = model_fit.forecast(steps = 1)[0]
+    # data1.NO2 = model_fit.forecast(steps = 1)[0]
+
+    smodel_o3 = sm.tsa.SARIMAX(df['O3'], order=my_order,seasonal_order=my_seasonal_order,trend= 't')
+    result_seasonal_o3 = smodel_o3.fit()
+    predictions_seasonal_o3 = result_seasonal_o3.predict(steps = 1)
+    # predictions_seasonal_o3 = result_seasonal_o3.forecast(steps = 1)[0]
+    print('o3        : ',predictions_seasonal_o3.iloc[-1])
+    data1.O3 = predictions_seasonal_o3.iloc[-1]
+    # data.O3 = predictions_seasonal_o3.iloc[-1]
+    # data1.O3 = 20
+
+    # model = ARIMA(df_train['O3'],order = (2,1,2))
+    # model_fit = model.fit()
+    # # data1.O3 = model_fit.forecast(steps = 1)[0]
+    # # future_datest_df_new['O3'] = model_fit.forecast(steps = 1)[0]
+    # data1.O3 = 80
+
+
+    smodel_pm10 = sm.tsa.SARIMAX(df['PM10'], order=my_order,seasonal_order=my_seasonal_order,trend= 't')
+    result_seasonal_pm10 = smodel_pm10.fit()
+    predictions_seasonal_pm10 = result_seasonal_pm10.predict(steps = 1)
+    # predictions_seasonal_pm10 = result_seasonal_pm10.forecast(steps = 1)[0]
+    print('pm10       : ',predictions_seasonal_pm10.iloc[-1])
+    data1.PM10 = predictions_seasonal_pm10.iloc[-1]
+    # data.PM10 = predictions_seasonal_pm10.iloc[-1]
+    # data1.PM10 = 20
+
+    # model = ARIMA(df_train['PM10'],order = (2,1,2))
+    # model_fit = model.fit()
+    # data1.PM10 = model_fit.forecast(steps = 1)[0]
+    # future_datest_df_new['PM10'] = model_fit.forecast(steps = 1)[0]
+    # data1.PM10 = model_fit.forecast(steps = 1)[0]
+
+
+    # df = pd.concat([df,future_datest_df_new])
+    # print(df.tail())
+    # data1.Date = future_dates[1].strftime('%Y-%m-%d')
+    data1.Date = pd.to_datetime(future_dates[1]).date()
+    # data.Date = pd.to_datetime(future_dates[1]).date()
+    data1.save()
+    # data.save()
+    # print(future_dates[1].strftime('%Y-%m-%d'))
+    # print(future_datest_df_new['AQI'])
+    # return df
+
+
+
+
+d1 = df['AQI'][-30:].tolist()
+d2 = df.index[-30:].strftime('%Y-%m-%d').tolist()
+
+d3 = df['SO2'][-30:].tolist()
+d4 = df['NO2'][-30:].tolist()
+d5 = df['O3'][-30:].tolist()
+
+print(df['Date'])
+print(df.tail())

main/Scraper.py

@@ -0,0 +1,50 @@
+import requests
+from datetime import datetime
+import pandas as pd
+from .models import  Data,Data_Predicted
+
+data = Data()
+data1 = Data_Predicted()
+
+
+
+def scrap():
+    url = "https://api.ambeedata.com/latest/by-city"
+    querystring = {"city":"Delhi"}
+    headers = {
+        'x-api-key': "mqoZ9hQjIi3ZqydD8W3lU7fGnXJ5ndJI44sAUusN",
+        'Content-type': "application/json"
+        }
+    response = requests.request("GET", url, headers=headers, params=querystring)
+
+    r = response.json()
+
+    data.NO2 = r['stations'][0]['NO2']
+    data.O3 = r['stations'][0]['OZONE']
+    data.SO2 = r['stations'][0]['SO2']
+    data.PM10 = r['stations'][0]['PM10']
+    data.AQI = r['stations'][0]['AQI']
+
+    x = pd.to_datetime(r['stations'][0]['updatedAt'])
+    data.Date = x.date()
+    data.save()
+
+    print('no2',r['stations'][0]['NO2'])
+    print('o3',r['stations'][0]['OZONE'])
+    print('so2',r['stations'][0]['SO2'])
+    print('pm10',r['stations'][0]['PM10'])
+    print('aqi',r['stations'][0]['AQI'])
+    print('date',r['stations'][0]['updatedAt'])
+    print('date type',type(r['stations'][0]['updatedAt']))
+
+
+
+    print(type(response))
+    print(type(r))
+
+    print(r['message'])
+    print(r['stations'][0]['OZONE'])
+
+    print(response.text)
+
+# scrap()

main/apscheduler/data_update_autojob.py

@@ -0,0 +1,12 @@
+# from apscheduler.schedulers.background import BackgroundScheduler
+
+# from . import Ml_model as ml
+# from .models import Test , Data
+
+# data = Data()
+# def start():
+#     scheduler = BackgroundScheduler()
+#     scheduler.add_job(ml.fun(ml.df),"interval",minutes = 1,id = "data_001")
+#     scheduler.start()
+    # Test.objects.create(name = "test")
+

main/cron.py

@@ -9,7 +9,7 @@ def my_scheduled_job():
     # data.AQI = 2
     # data.save()
     ml.fun(ml.df)
-    sc.scrap()
+    # sc.scrap()
 
 
     # ml.f_s(ml.df)

main/models.py

@@ -0,0 +1,22 @@
+from django.db import models
+
+# Create your models here.
+class Data(models.Model):
+    Date = models.DateField()
+    PM10 = models.IntegerField()
+    NO2 = models.IntegerField()
+    SO2 = models.IntegerField()
+    O3 = models.IntegerField()
+    AQI = models.IntegerField()
+
+
+class Data_Predicted(models.Model):
+    Date = models.DateField()
+    PM10 = models.IntegerField()
+    NO2 = models.IntegerField()
+    SO2 = models.IntegerField()
+    O3 = models.IntegerField()
+    AQI = models.IntegerField()
+
+class Update(models.Model):
+    dat_up = models.IntegerField()

main/resources.py

@@ -0,0 +1,6 @@
+from import_export import resources
+from .models import Data
+
+class DataResources(resources.ModelResource):
+    class Meta:
+        model = Data

main/sc.py

@@ -0,0 +1,17 @@
+import requests
+from datetime import datetime
+
+aq = []
+def scrap():
+    url = "http://vc8006.pythonanywhere.com/api/"
+    response = requests.request("GET", url)
+
+    r = response.json()
+    for i in range(1,31):
+        aq.append(r[-i]['AQI'])
+        # print(r[-i])
+
+    # print(response.text)
+    print(aq)
+
+scrap()

main/serializers.py

@@ -0,0 +1,7 @@
+from rest_framework import fields, serializers
+from .models import Data_Predicted
+
+class DataPredSerializers(serializers.ModelSerializer):
+    class Meta:
+        fields = ('Date','PM10','NO2','SO2','O3','AQI')
+        model = Data_Predicted