9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [1]:  import pandas as pd

import matplotlib.pyplot as plt
import seaborn as sns
import datetime as dt

In [2]:  df = pd.read_csv("uber.csv")
df.head()

Out[2]:
Unnamed:
key fare_amount pickup_datetime pickup_longitude pick
0

2015-05-07 2015-05-07
0 24238194 7.5 -73.999817
19:52:06.0000003 19:52:06 UTC

2009-07-17 2009-07-17
1 27835199 7.7 -73.994355
20:04:56.0000002 20:04:56 UTC

2009-08-24 2009-08-24
2 44984355 12.9 -74.005043
21:45:00.00000061 21:45:00 UTC

2009-06-26 2009-06-26
3 25894730 5.3 -73.976124
08:22:21.0000001 08:22:21 UTC

2014-08-28 2014-08-28
4 17610152 16.0 -73.925023
17:47:00.000000188 17:47:00 UTC

In [3]:  df.drop(columns=['Unnamed: 0','key'],inplace=True)

In [4]:  df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 200000 entries, 0 to 199999
Data columns (total 7 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 fare_amount 200000 non-null float64
1 pickup_datetime 200000 non-null object
2 pickup_longitude 200000 non-null float64
3 pickup_latitude 200000 non-null float64
4 dropoff_longitude 199999 non-null float64
5 dropoff_latitude 199999 non-null float64
6 passenger_count 200000 non-null int64
dtypes: float64(5), int64(1), object(1)
memory usage: 10.7+ MB

In [5]:  df.dropna(how='any',inplace=True)

localhost:8888/notebooks/Uber ml1.ipynb 1/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [6]:  df.isnull().sum()

Out[6]: fare_amount 0
pickup_datetime 0
pickup_longitude 0
pickup_latitude 0
dropoff_longitude 0
dropoff_latitude 0
passenger_count 0
dtype: int64

Boxplots

In [7]:  for col in df.select_dtypes(exclude=['object']):

plt.figure()
sns.boxplot(data=df,x=col)

In [8]:  df = df[
(df.pickup_latitude > -90) & (df.pickup_latitude < 90) &
(df.dropoff_latitude > -90) & (df.dropoff_latitude < 90) &
(df.pickup_longitude > -180) & (df.pickup_longitude < 180) &
(df.dropoff_longitude > -180) & (df.dropoff_longitude < 180) &
(df.fare_amount > 0) & (df.passenger_count > 0) & (df.passenger_co
]

localhost:8888/notebooks/Uber ml1.ipynb 2/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [9]:  from math import cos, asin, sqrt, pi

import numpy as np
​
def distance(lat_1,lon_1,lat_2,lon_2):
# lat1 = row.pickup_latitude
# lon1 = row.pickup_longitude
# lat2 = row.dropoff_latitude
# lon2 = row.dropoff_longitude
lon_1, lon_2, lat_1, lat_2 = map(np.radians, [lon_1, lon_2, lat_1,

diff_lon = lon_2 - lon_1

diff_lat = lat_2 - lat_1

​
km = 2 * 6371 * np.arcsin(np.sqrt(np.sin(diff_lat/2.0)**2 + np.cos

return km

In [10]:  temp = distance(df['pickup_latitude'],df['pickup_longitude'],df['dropof

temp.head()

Out[10]: 0 1.683323
1 2.457590
2 5.036377
3 1.661683
4 4.475450
dtype: float64

In [11]:  df_new = df.copy()

df_new['Distance'] = temp
df = df_new
df.head()

Out[11]:
fare_amount pickup_datetime pickup_longitude pickup_latitude dropoff_longitude d

2015-05-07
0 7.5 -73.999817 40.738354 -73.999512
19:52:06 UTC

2009-07-17
1 7.7 -73.994355 40.728225 -73.994710
20:04:56 UTC

2009-08-24
2 12.9 -74.005043 40.740770 -73.962565
21:45:00 UTC

2009-06-26
3 5.3 -73.976124 40.790844 -73.965316
08:22:21 UTC

2014-08-28
4 16.0 -73.925023 40.744085 -73.973082
17:47:00 UTC

localhost:8888/notebooks/Uber ml1.ipynb 3/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [12]:  sns.boxplot(data=df,x='Distance')

Out[12]: <Axes: xlabel='Distance'>

In [13]:  df = df[(df['Distance'] < 200) & (df['Distance'] > 0)]

In [14]:  df['pickup_datetime'] = pd.to_datetime(df['pickup_datetime'])

C:\Users\HP\AppData\Local\Temp\ipykernel_16404\1295461447.py:1: Setti
ngWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/panda

s-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
(https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.htm
l#returning-a-view-versus-a-copy)
df['pickup_datetime'] = pd.to_datetime(df['pickup_datetime'])

localhost:8888/notebooks/Uber ml1.ipynb 4/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [15]:  df['week_day'] = df['pickup_datetime'].dt.day_name()

df['Year'] = df['pickup_datetime'].dt.year
df['Month'] = df['pickup_datetime'].dt.month
df['Hour'] = df['pickup_datetime'].dt.hour

C:\Users\HP\AppData\Local\Temp\ipykernel_16404\2592915223.py:1: Setti
ngWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/panda

s-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
(https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.htm
l#returning-a-view-versus-a-copy)
df['week_day'] = df['pickup_datetime'].dt.day_name()
C:\Users\HP\AppData\Local\Temp\ipykernel_16404\2592915223.py:2: Setti
ngWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/panda

s-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
(https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.htm
l#returning-a-view-versus-a-copy)
df['Year'] = df['pickup_datetime'].dt.year
C:\Users\HP\AppData\Local\Temp\ipykernel_16404\2592915223.py:3: Setti
ngWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/panda

s-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
(https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.htm
l#returning-a-view-versus-a-copy)
df['Month'] = df['pickup_datetime'].dt.month
C:\Users\HP\AppData\Local\Temp\ipykernel_16404\2592915223.py:4: Setti
ngWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/panda

s-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
(https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.htm
l#returning-a-view-versus-a-copy)
df['Hour'] = df['pickup_datetime'].dt.hour

In [16]:  df.drop(columns=['pickup_datetime','pickup_latitude','pickup_longitude'

C:\Users\HP\AppData\Local\Temp\ipykernel_16404\3782303944.py:1: Setti
ngWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/panda

s-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
(https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.htm
l#returning-a-view-versus-a-copy)
df.drop(columns=['pickup_datetime','pickup_latitude','pickup_longit
ude','dropoff_latitude','dropoff_longitude'],inplace=True)

localhost:8888/notebooks/Uber ml1.ipynb 5/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [17]:  df.head()

Out[17]:
fare_amount passenger_count Distance week_day Year Month Hour

0 7.5 1 1.683323 Thursday 2015 5 19

1 7.7 1 2.457590 Friday 2009 7 20

2 12.9 1 5.036377 Monday 2009 8 21

3 5.3 3 1.661683 Friday 2009 6 8

4 16.0 5 4.475450 Thursday 2014 8 17

localhost:8888/notebooks/Uber ml1.ipynb 6/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [18]:  temp = df.copy()

​
def convert_week_day(day):
if day in ['Monday','Tuesday','Wednesday','Thursday']:
return 0 # Weekday
return 1 # Weekend
​
def convert_hour(hour):
if 5 <= hour <= 12:
return 1
elif 12 < hour <= 17:
return 2
elif 17 < hour < 24:
return 3
return 0
​
df['week_day'] = temp['week_day'].apply(convert_week_day)
df['Hour'] = temp['Hour'].apply(convert_hour)
df.head()

C:\Users\HP\AppData\Local\Temp\ipykernel_16404\3260682206.py:17: Sett
ingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/panda

s-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
(https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.htm
l#returning-a-view-versus-a-copy)
df['week_day'] = temp['week_day'].apply(convert_week_day)
C:\Users\HP\AppData\Local\Temp\ipykernel_16404\3260682206.py:18: Sett
ingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/panda

s-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
(https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.htm
l#returning-a-view-versus-a-copy)
df['Hour'] = temp['Hour'].apply(convert_hour)

Out[18]:
fare_amount passenger_count Distance week_day Year Month Hour

0 7.5 1 1.683323 0 2015 5 3

1 7.7 1 2.457590 1 2009 7 3

2 12.9 1 5.036377 0 2009 8 3

3 5.3 3 1.661683 1 2009 6 1

4 16.0 5 4.475450 0 2014 8 2

localhost:8888/notebooks/Uber ml1.ipynb 7/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [19]:  df.corr()

Out[19]:
fare_amount passenger_count Distance week_day Year Mo

fare_amount 1.000000 0.011884 0.778667 0.002305 0.120430 0.024

passenger_count 0.011884 1.000000 0.005112 0.035882 0.005339 0.008

Distance 0.778667 0.005112 1.000000 0.014518 0.018617 0.007

week_day 0.002305 0.035882 0.014518 1.000000 0.006910 -0.007

Year 0.120430 0.005339 0.018617 0.006910 1.000000 -0.115

Month 0.024120 0.008818 0.007373 -0.007328 -0.115182 1.000

Hour -0.021078 0.013572 -0.022691 -0.078129 0.001131 -0.005

In [20]:  sns.scatterplot(y=df['fare_amount'],x=df['Distance'])

Out[20]: <Axes: xlabel='Distance', ylabel='fare_amount'>

In [21]:  from sklearn.preprocessing import StandardScaler

x = df[['Distance']].values
y = df['fare_amount'].values.reshape(-1,1)

In [22]:  from sklearn.model_selection import train_test_split

x_train, x_test, y_train,y_test = train_test_split(x,y,random_state=10)

localhost:8888/notebooks/Uber ml1.ipynb 8/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

In [23]:  std_x = StandardScaler()

x_train = std_x.fit_transform(x_train)

In [24]:  x_test = std_x.transform(x_test)

In [25]:  std_y = StandardScaler()

y_train = std_y.fit_transform(y_train)

In [26]:  y_test = std_y.transform(y_test)

In [27]:  from sklearn.metrics import mean_squared_error,r2_score, mean_absolute_

def fit_predict(model):
model.fit(x_train,y_train.ravel())
y_pred = model.predict(x_test)
r_squared = r2_score(y_test,y_pred)
RMSE = mean_squared_error(y_test, y_pred,squared=False)
MAE = mean_absolute_error(y_test,y_pred)
print('R-squared: ', r_squared)
print('RMSE: ', RMSE)
print("MAE: ",MAE)

In [28]:  from sklearn.linear_model import LinearRegression

In [29]:  fit_predict(LinearRegression())

R-squared: 0.6041167920841171
RMSE: 0.6290054895695945
MAE: 0.27552329590959806

C:\Users\HP\AppData\Roaming\Python\Python311\site-packages\sklearn\me
trics\_regression.py:483: FutureWarning: 'squared' is deprecated in v
ersion 1.4 and will be removed in 1.6. To calculate the root mean squ
ared error, use the function'root_mean_squared_error'.
warnings.warn(

In [30]:  from sklearn.ensemble import RandomForestRegressor

fit_predict(RandomForestRegressor())

R-squared: 0.6522221648884474
RMSE: 0.5895516309915084
MAE: 0.2918258149086775

C:\Users\HP\AppData\Roaming\Python\Python311\site-packages\sklearn\me
trics\_regression.py:483: FutureWarning: 'squared' is deprecated in v
ersion 1.4 and will be removed in 1.6. To calculate the root mean squ
ared error, use the function'root_mean_squared_error'.
warnings.warn(

localhost:8888/notebooks/Uber ml1.ipynb 9/10

9/17/24, 8:10 PM Uber ml1 - Jupyter Notebook

localhost:8888/notebooks/Uber ml1.ipynb 10/10