Commit Live PR

q01_load_data/build.py

@@ -1,7 +1,16 @@
+# %load q01_load_data/build.py
 import pandas as pd
 import numpy as np
 from sklearn.model_selection import train_test_split
 
+def q01_load_data(path):
+    tss=pd.read_csv(path)
+    tss['Datetime']=pd.to_datetime(tss['Datetime'])
+    return tss.shape, tss
+
+
 
 
 
+
+

q02_data_splitter/build.py

@@ -1,7 +1,19 @@
+# %load q02_data_splitter/build.py
 import pandas as pd
 import numpy as np
 from sklearn.model_selection import TimeSeriesSplit
 from greyatomlib.time_series_day_02_project.q01_load_data.build import q01_load_data
 
+def q02_data_splitter(path):
+    seed=9
+    shape,df = q01_load_data(path)
+    tssf= TimeSeriesSplit(n_splits=3)
+    trainl=()
+    validl=()
+    for train_index,valid_index in tssf.split(df):
+        trainl=trainl+ tuple(train_index)
+        validl=validl + tuple(valid_index)
+    return [[trainl,trainl],[validl,validl]]
+
+
 
-

q03_time_plot/build.py

@@ -1,7 +1,18 @@
+# %load q03_time_plot/build.py
 import pandas as pd
 import numpy as np
 import matplotlib.pyplot as plt
+
 from greyatomlib.time_series_day_02_project.q01_load_data.build import q01_load_data
 plt.switch_backend('agg')
+def q03_time_plot(path):
+    shp,df=q01_load_data(path)
+    plt.figure(figsize=(16, 6))
+    plt.plot(df['Datetime'], df['Demand'])
+    plt.xlabel('Time')
+    plt.ylabel('Demand')
+    plt.title('SateTime vs Demand in Australia')
+    plt.show()
 
+
 

q04_boxplot/build.py

@@ -1,7 +1,18 @@
+# %load q04_boxplot/build.py
 import pandas as pd
 import numpy as np
+import seaborn as sns
 import matplotlib.pyplot as plt
 from greyatomlib.time_series_day_02_project.q01_load_data.build import q01_load_data
 plt.switch_backend('agg')
 
-
+def q04_boxplot(path):
+
+    shp,df=q01_load_data(path)
+    plt.figure(figsize=(16, 7))
+    sns.factorplot(x='WorkDay', y='Demand', data=df, kind='box', size=8, aspect=float(16/7))
+    plt.xlabel('Workday')
+    plt.ylabel('Demand')
+    plt.title('Each workday demand in Australia')
+
+

q05_feature_engineering/build.py

@@ -1,3 +1,4 @@
+# %load q05_feature_engineering/build.py
 import pandas as pd
 import numpy as np
 import matplotlib.pyplot as plt
@@ -6,4 +7,12 @@
 
 
 
+def q05_feature_engineering(path):
+    tss,df=q01_load_data(path)
+    p_corf=np.corrcoef(df['Temperature'], df['Demand'])
+    plt.scatter(df['Temperature'], df['Demand'])
+    plt.show()
+
+#path='data/elecdemand.csv'
+#q05_feature_engineering(path)
 

q05_feature_engineering_part2/build.py

@@ -1,8 +1,23 @@
+# %load q05_feature_engineering_part2/build.py
 import pandas as pd
 import numpy as np
 import matplotlib.pyplot as plt
-from greyatomlib.time_series_day_02_project.q01_load_data.build import q01_load_data 
+from greyatomlib.time_series_day_02_project.q01_load_data.build import q01_load_data
 plt.switch_backend('agg')
 
+def q05_feature_engineering_part2(path):
+    shape, data = q01_load_data(path)
+    'write your solution here'
+    data['hour'] = data['Datetime'].dt.hour
+
+    plt.figure(figsize=(16, 6))
+
+    hours = []
+    for i in range(24):
+        one = data[data['hour'] == i]['Demand'].values
+        hours.append(one)
+    plt.boxplot(hours, labels=[str(i) for i in range(24)])
+    plt.xlabel('Hour')
+    plt.ylabel('Demand')
+    plt.title('Change in Electricity demand wrt to Hour')    
 
-

q05_feature_engineering_part3/build.py

@@ -1,8 +1,18 @@
+# %load q05_feature_engineering_part3/build.py
 import pandas as pd
 import numpy as np
 import matplotlib.pyplot as plt
+import seaborn as sns
 from greyatomlib.time_series_day_02_project.q01_load_data.build import q01_load_data
 plt.switch_backend('agg')
-
-
+def q05_feature_engineering_part3(path):
+    tss,df=q01_load_data(path)
+    df.info()
+    df['hour']=df['Datetime'].dt.hour
+    df['month']=df['Datetime'].dt.strftime('%b')
 
+    sns.factorplot(x='month', y='Demand', data=df, kind='box', order=['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug','Sep', 'Oct', 'Nov', 'Dec'], size=8, aspect=float(16/7))
+
+#path='data/elecdemand.csv'    
+#q05_feature_engineering_part3(path)
+

q05_feature_engineering_part4/build.py

@@ -1,9 +1,18 @@
+# %load q05_feature_engineering_part2/build.py
 import pandas as pd
 import numpy as np
-from sklearn.model_selection import train_test_split
 import matplotlib.pyplot as plt
 from greyatomlib.time_series_day_02_project.q01_load_data.build import q01_load_data
-plt.switch_backend('agg')
 
-def q05_feature_engineering_part4():    
-
+
+def q05_feature_engineering_part4(path):
+    shape, data = q01_load_data(path)
+    data['hour'] = data['Datetime'].dt.hour
+    data['month'] = data['Datetime'].dt.strftime('%b')
+    data['Peakhours']=data['hour'].apply(lambda x : 1 if x in range(6,20) else 0)
+    data['Peakmonths']=data['month'].apply(lambda x : 1 if x in ['Feb', 'May', 'Jun', 'Jul', 'Aug'] else 0)
+    return data
+
+#path='data/elecdemand.csv'
+#q05_feature_engineering_part4(path)
+

q06_linear_regression/build.py

@@ -1,12 +1,34 @@
+# %load q06_linear_regression/build.py
 import pandas as pd
 import numpy as np
 import math
 from sklearn.linear_model import LinearRegression
 from sklearn.metrics import mean_squared_error
+
+from greyatomlib.time_series_day_02_project.q01_load_data.build import q01_load_data
 from greyatomlib.time_series_day_02_project.q05_feature_engineering_part4.build import q05_feature_engineering_part4
 from greyatomlib.time_series_day_02_project.q02_data_splitter.build import q02_data_splitter
 
-fe =  ["WorkDay", "Peakhours", "Peakmonths"]
-
+fe =  ['WorkDay', 'Peakhours', 'Peakmonths']
 
+def q06_linear_regression(path,columns = fe, random_state =9):
+    np.random.seed(random_state)
+    data = q05_feature_engineering_part4(path)
+    splits = q02_data_splitter(path)
+    'write your solution here'
 
+    rmse = []
+    for i in splits:
+        train = i[0]
+        valid = i[1]
+        x_train, y_train = data[fe].values[train], data['Demand'].values[train]
+        x_valid, y_valid = data[fe].values[valid], data['Demand'].values[valid]
+        model = LinearRegression()
+        model.fit(x_train, y_train)
+        pred = model.predict(x_valid)
+        measure = math.pow(mean_squared_error(y_valid, pred), 0.5)
+        rmse.append(measure)
+    return np.mean(rmse)
+
+
+

q07_randomforest_regressor/build.py

@@ -1,3 +1,4 @@
+# %load q07_randomforest_regressor/build.py
 import pandas as pd
 import numpy as np
 import math
@@ -6,7 +7,28 @@
 from greyatomlib.time_series_day_02_project.q05_feature_engineering_part4.build import q05_feature_engineering_part4
 from greyatomlib.time_series_day_02_project.q02_data_splitter.build import q02_data_splitter
 
-fe =  ["WorkDay", "Peakhours", "Peakmonths"]
+fe =  ['WorkDay', 'Peakhours', 'Peakmonths']
 
+def q07_randomforest_regressor(path,columns = fe, random_state =9):
+    np.random.seed(random_state)
+    data = q05_feature_engineering_part4(path)
+    splits = q02_data_splitter(path)
+    rmse = []
 
+    for i in splits:
+        train = i[0]
+        valid = i[1]
+        x_train, y_train = data[fe].values[train], data['Demand'].values[train]
+        x_valid, y_valid = data[fe].values[valid], data['Demand'].values[valid]
+        model = RandomForestRegressor( n_estimators=50, min_samples_leaf=30, random_state=10)
+        model.fit(x_train, y_train)
+        pred = model.predict(x_valid)
+        measure = math.pow(mean_squared_error(y_valid, pred), 0.5)
+        rmse.append(measure)
+    return np.mean(rmse)
 
+
+
+#path=   'data/elecdemand.csv' 
+#q07_randomforest_regressor(path)
+

q08_gradientboosting_regressor/build.py

@@ -1,3 +1,4 @@
+# %load q08_gradientboosting_regressor/build.py
 import pandas as pd
 import numpy as np
 import math
@@ -6,5 +7,31 @@
 from greyatomlib.time_series_day_02_project.q05_feature_engineering_part4.build import q05_feature_engineering_part4
 from greyatomlib.time_series_day_02_project.q02_data_splitter.build import q02_data_splitter
 
-fe =  ["WorkDay", "Peakhours", "Peakmonths"]
+fe =  ['WorkDay', 'Peakhours', 'Peakmonths']
+
+
+def q08_gradientboosting_regressor(path,columns = fe, random_state =9):
+    np.random.seed(random_state)
+    data = q05_feature_engineering_part4(path)
+    splits = q02_data_splitter(path)
+    rmse = []
+
+    for i in splits:
+        train = i[0]
+        valid = i[1]
+        x_train, y_train = data[fe].values[train], data['Demand'].values[train]
+        x_valid, y_valid = data[fe].values[valid], data['Demand'].values[valid]
+        model = GradientBoostingRegressor( n_estimators=200, min_samples_leaf=10, learning_rate=0.01, random_state=random_state)
+        model.fit(x_train, y_train)
+        pred = model.predict(x_valid)
+        measure = math.pow(mean_squared_error(y_valid, pred), 0.5)
+        rmse.append(measure)
+    return np.mean(rmse)
+
+
+
+#path=   'data/elecdemand.csv' 
+#q08_gradientboosting_regressor(path)
 
+
+