Car trips data log 확인

데이터: https://www.kaggle.com/datasets/vitorrf/cartripsdatamining

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import sklearn
import warnings
warnings.filterwarnings('ignore')

df_list = []

for i in range(10, 15):
    file_name = f'/kaggle/input/cartripsdatamining/Processed Data/fileID{i}_ProcessedTripData.csv'

    df = pd.read_csv(
        file_name,
        header= None,
    )
    
    df.columns=['Time','Vehicle Speed','SHIFT','Engine Load','Total Acceleration',
            'Engine RPM','Pitch','Lateral Acceleration','Passenger Count',
            'Car Load','AC Status','Window Opening','Radio Volume','Rain Intensity',
            'Visibility','Driver Wellbeing','Driver Rush']
    
    df_list.append(df)

df.head()

	Time	Vehicle Speed	SHIFT	Engine Load	Total Acceleration	Engine RPM	Pitch	Lateral Acceleration	Passenger Count	Car Load	AC Status	Window Opening	Radio Volume	Rain Intensity	Visibility	Driver Wellbeing	Driver Rush
0	0.019	0.0	0	0.0	0.0	0.0	-0.050212	-0.0	3	3	0	0	2	0	10	6	2
1	0.026	0.0	0	0.0	0.0	0.0	-0.051910	-0.0	3	3	0	0	2	0	10	6	2
2	0.037	0.0	0	0.0	0.0	0.0	-0.053624	-0.0	3	3	0	0	2	0	10	6	2
3	0.048	0.0	0	0.0	0.0	0.0	-0.055352	-0.0	3	3	0	0	2	0	10	6	2
4	0.056	0.0	0	0.0	0.0	0.0	-0.057097	-0.0	3	3	0	0	2	0	10	6	2

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 118551 entries, 0 to 118550
Data columns (total 17 columns):
 #   Column                Non-Null Count   Dtype  
---  ------                --------------   -----  
 0   Time                  118551 non-null  float64
 1   Vehicle Speed         118551 non-null  float64
 2   SHIFT                 118551 non-null  int64  
 3   Engine Load           118551 non-null  float64
 4   Total Acceleration    118551 non-null  float64
 5   Engine RPM            118551 non-null  float64
 6   Pitch                 118551 non-null  float64
 7   Lateral Acceleration  118551 non-null  float64
 8   Passenger Count       118551 non-null  int64  
 9   Car Load              118551 non-null  int64  
 10  AC Status             118551 non-null  int64  
 11  Window Opening        118551 non-null  int64  
 12  Radio Volume          118551 non-null  int64  
 13  Rain Intensity        118551 non-null  int64  
 14  Visibility            118551 non-null  int64  
 15  Driver Wellbeing      118551 non-null  int64  
 16  Driver Rush           118551 non-null  int64  
dtypes: float64(7), int64(10)
memory usage: 15.4 MB

cols = ['Vehicle Speed','SHIFT','Engine Load','Total Acceleration',
            'Engine RPM','Pitch','Lateral Acceleration']

fig, axs = plt.subplots(
    nrows=len(cols), 
    figsize=(8, 2*len(cols)), 
    sharex='col',
    sharey='row',
)

for ax, col in zip(axs, cols):
    sns.lineplot(df[:2000], x='Time', y=col, ax=ax)
    
plt.tight_layout()
plt.show()

from scipy.integrate import cumtrapz

df['Distance'] = cumtrapz(df['Vehicle Speed'], df['Time'], initial=0)

cols = ['Vehicle Speed','Distance']

fig, axs = plt.subplots(
    nrows=len(cols), 
    figsize=(8, 2*len(cols)), 
    sharex='col',
    sharey='row',
)

for ax, col in zip(axs, cols):
    sns.lineplot(df[:4000], x='Time', y=col, ax=ax)
    
plt.tight_layout()
plt.show()

df['Abs Acceleration'] = df['Total Acceleration'].abs()

sns.lineplot(df[:4000], x='Distance', y='Abs Acceleration')
plt.fill_between(df[:4000]['Distance'], df[:4000]['Abs Acceleration'], color='skyblue', alpha=0.3)
plt.show()

from scipy.integrate import trapz

work = trapz(df['Abs Acceleration'], df['Distance'])
work

5260.302733112259

df['Work'] = cumtrapz(df['Abs Acceleration'], df['Distance'], initial=0)

def get_dist_col(df):
    df['Distance'] = cumtrapz(df['Vehicle Speed'], df['Time'], initial=0)
    return df

def get_work_col(df, to_abs=True):
    if to_abs:
        df['Work'] = cumtrapz(df['Total Acceleration'].abs(), df['Distance'], initial=0)
    else:
        df['Work'] = cumtrapz(df['Total Acceleration'], df['Distance'], initial=0)
    return df

df_list = [get_dist_col(df) for df in df_list]
df_list = [get_work_col(df) for df in df_list]

df = df_list[2]

cols = ['Vehicle Speed','SHIFT','Engine Load','Total Acceleration',
            'Engine RPM','Pitch','Lateral Acceleration']

fig, axs = plt.subplots(
    nrows=len(cols), 
    figsize=(8, 2*len(cols)), 
    sharex='col',
    sharey='row',
)

for ax, col in zip(axs, cols):
    sns.lineplot(df[:2000], x='Time', y=col, ax=ax)
    
plt.tight_layout()
plt.show()