import numpy as np
import pandas as pd
import matplotlib
import matplotlib.pyplot as plt
from matplotlib.path import Path
from matplotlib.patches import PathPatch
import seaborn as sns
from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LogisticRegression
from sklearn import metrics

Masyvą sudaro 1992-2020 m. auto pardavimų duomenys Delyje

1. Hipotezė. Daugiau kaip 50 K km nuvažiavę bei turintys "mechaninę" pavarų dėžę automobiliai yra pigesni. (Nepasitvirtino).
2. Hipotezė. Individualūs savininkai automobilius parduoda pigiau (mažiau reklamos, nesuteikia garantijos, patikimumas ir kt.). (Visiškai pasitvirtino).
3. Hipotezė. Norint įsigyti automobilį nuvažiavusį ne daugiau 50 K km ir ne senesnį kaip 3 metų amžiaus, dažniausiai reikia "pasiruošti" bent 10 k EUR. (Dažniausiai taip).
4. Hipotezė. Jei pirkėjas turi iki 3 k EUR, reikėtų tikėtis automobilio su "mechanine" pavarų dėže ir nuvažiavusio bent 100 k km. (Visiškai nepasitvirtino, bent jau Delio atveju).
5. Prognozė. Vertinant automobilio ridą, kokia turėtų būti jo kaina.

car_deho_india = pd.read_csv('car_deho_india_original.csv')
car = car_deho_india.dropna()
car

	name	year	selling_price	km_driven	fuel	seller_type	transmission	owner
0	Maruti 800 AC	2007	60000	70000	Petrol	Individual	Manual	First Owner
1	Maruti Wagon R LXI Minor	2007	135000	50000	Petrol	Individual	Manual	First Owner
2	Hyundai Verna 1.6 SX	2012	600000	100000	Diesel	Individual	Manual	First Owner
3	Datsun RediGO T Option	2017	250000	46000	Petrol	Individual	Manual	First Owner
4	Honda Amaze VX i-DTEC	2014	450000	141000	Diesel	Individual	Manual	Second Owner
...	...	...	...	...	...	...	...	...
4335	Hyundai i20 Magna 1.4 CRDi (Diesel)	2014	409999	80000	Diesel	Individual	Manual	Second Owner
4336	Hyundai i20 Magna 1.4 CRDi	2014	409999	80000	Diesel	Individual	Manual	Second Owner
4337	Maruti 800 AC BSIII	2009	110000	83000	Petrol	Individual	Manual	Second Owner
4338	Hyundai Creta 1.6 CRDi SX Option	2016	865000	90000	Diesel	Individual	Manual	First Owner
4339	Renault KWID RXT	2016	225000	40000	Petrol	Individual	Manual	First Owner

4340 rows × 8 columns

car.nunique()

name             1491
year               27
selling_price     445
km_driven         770
fuel                5
seller_type         3
transmission        2
owner               5
make               29
dtype: int64

car['make'] = car['name'].str.split(' ').str[0]
car

	name	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make
0	Maruti 800 AC	2007	60000	70000	Petrol	Individual	Manual	First Owner	Maruti
1	Maruti Wagon R LXI Minor	2007	135000	50000	Petrol	Individual	Manual	First Owner	Maruti
2	Hyundai Verna 1.6 SX	2012	600000	100000	Diesel	Individual	Manual	First Owner	Hyundai
3	Datsun RediGO T Option	2017	250000	46000	Petrol	Individual	Manual	First Owner	Datsun
4	Honda Amaze VX i-DTEC	2014	450000	141000	Diesel	Individual	Manual	Second Owner	Honda
...	...	...	...	...	...	...	...	...	...
4335	Hyundai i20 Magna 1.4 CRDi (Diesel)	2014	409999	80000	Diesel	Individual	Manual	Second Owner	Hyundai
4336	Hyundai i20 Magna 1.4 CRDi	2014	409999	80000	Diesel	Individual	Manual	Second Owner	Hyundai
4337	Maruti 800 AC BSIII	2009	110000	83000	Petrol	Individual	Manual	Second Owner	Maruti
4338	Hyundai Creta 1.6 CRDi SX Option	2016	865000	90000	Diesel	Individual	Manual	First Owner	Hyundai
4339	Renault KWID RXT	2016	225000	40000	Petrol	Individual	Manual	First Owner	Renault

4340 rows × 9 columns

car.groupby('name')['name'].count()

name
Ambassador CLASSIC 1500 DSL AC          2
Ambassador Classic 2000 Dsz             1
Ambassador Grand 1800 ISZ MPFI PW CL    1
Audi A4 1.8 TFSI                        1
Audi A4 2.0 TDI                         1
                                       ..
Volkswagen Vento Petrol Highline AT     2
Volvo V40 D3 R Design                   1
Volvo XC 90 D5 Inscription BSIV         1
Volvo XC60 D3 Kinetic                   1
Volvo XC60 D5 Inscription               1
Name: name, Length: 1491, dtype: int64

car.drop('name', axis=1, inplace=True)
car

	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make
0	2007	60000	70000	Petrol	Individual	Manual	First Owner	Maruti
1	2007	135000	50000	Petrol	Individual	Manual	First Owner	Maruti
2	2012	600000	100000	Diesel	Individual	Manual	First Owner	Hyundai
3	2017	250000	46000	Petrol	Individual	Manual	First Owner	Datsun
4	2014	450000	141000	Diesel	Individual	Manual	Second Owner	Honda
...	...	...	...	...	...	...	...	...
4335	2014	409999	80000	Diesel	Individual	Manual	Second Owner	Hyundai
4336	2014	409999	80000	Diesel	Individual	Manual	Second Owner	Hyundai
4337	2009	110000	83000	Petrol	Individual	Manual	Second Owner	Maruti
4338	2016	865000	90000	Diesel	Individual	Manual	First Owner	Hyundai
4339	2016	225000	40000	Petrol	Individual	Manual	First Owner	Renault

4340 rows × 8 columns

car['price_EUR'] = (car['selling_price'] * 0.012).round(0)
car['price_EUR']

0         720.0
1        1620.0
2        7200.0
3        3000.0
4        5400.0
         ...   
4335     4920.0
4336     4920.0
4337     1320.0
4338    10380.0
4339     2700.0
Name: price_EUR, Length: 4340, dtype: float64

car['price_EUR'].sort_values()

2662       240.0
2495       264.0
2444       360.0
3206       420.0
2849       420.0
          ...   
1023     59400.0
2239     59400.0
3969     66000.0
89       97800.0
3872    106800.0
Name: price_EUR, Length: 4340, dtype: float64

def if_salygos(x):
    if x < 1000:
        return "Iki 1000"
    elif x >= 1000 and x < 2000:
        return "Iki 2000"
    elif x >= 2000 and x < 3000:
        return "Iki 3000"
    elif x >= 3000 and x < 4000:
        return "Iki 4000"
    elif x >= 4000 and x < 5000:
        return "Iki 5000"
    elif x >= 5000 and x < 7000:
        return "Iki 7000"
    elif x >= 7000 and x < 10000:
        return "Iki 10000"
    elif x >= 10000 and x < 20000:
        return "Iki 20000"
    else:
        return "Nuo 20000"

reiksmes_vektorizuoti = np.vectorize(if_salygos)
sugrupuoti = reiksmes_vektorizuoti(car['price_EUR'])

car['price_group'] = sugrupuoti
car

	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make	price_EUR	price_group
0	2007	60000	70000	Petrol	Individual	Manual	First Owner	Maruti	720.0	Iki 1000
1	2007	135000	50000	Petrol	Individual	Manual	First Owner	Maruti	1620.0	Iki 2000
2	2012	600000	100000	Diesel	Individual	Manual	First Owner	Hyundai	7200.0	Iki 10000
3	2017	250000	46000	Petrol	Individual	Manual	First Owner	Datsun	3000.0	Iki 4000
4	2014	450000	141000	Diesel	Individual	Manual	Second Owner	Honda	5400.0	Iki 7000
...	...	...	...	...	...	...	...	...	...	...
4335	2014	409999	80000	Diesel	Individual	Manual	Second Owner	Hyundai	4920.0	Iki 5000
4336	2014	409999	80000	Diesel	Individual	Manual	Second Owner	Hyundai	4920.0	Iki 5000
4337	2009	110000	83000	Petrol	Individual	Manual	Second Owner	Maruti	1320.0	Iki 2000
4338	2016	865000	90000	Diesel	Individual	Manual	First Owner	Hyundai	10380.0	Iki 20000
4339	2016	225000	40000	Petrol	Individual	Manual	First Owner	Renault	2700.0	Iki 3000

4340 rows × 10 columns

car.groupby('price_group').count()

	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make	price_EUR
price_group
Iki 1000	236	236	236	236	236	236	236	236	236
Iki 10000	656	656	656	656	656	656	656	656	656
Iki 2000	600	600	600	600	600	600	600	600	600
Iki 20000	388	388	388	388	388	388	388	388	388
Iki 3000	532	532	532	532	532	532	532	532	532
Iki 4000	652	652	652	652	652	652	652	652	652
Iki 5000	431	431	431	431	431	431	431	431	431
Iki 7000	706	706	706	706	706	706	706	706	706
Nuo 20000	139	139	139	139	139	139	139	139	139

Hipotezė 1

Daugiau kaip 50 K km nuvažiavę bei turintys "mechaninę" pavarų dėžę automobiliai yra pigesni.

(Nepasitvirtino)

car.head(1)

	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make	price_EUR	price_group
0	2007	60000	70000	Petrol	Individual	Manual	First Owner	Maruti	720.0	Iki 1000

car[car['km_driven'] >= 50000].groupby(['price_group', 'transmission'])['price_EUR'].count()

price_group  transmission
Iki 1000     Automatic         1
             Manual          192
Iki 10000    Automatic        22
             Manual          282
Iki 2000     Automatic        11
             Manual          514
Iki 20000    Automatic        62
             Manual          105
Iki 3000     Automatic         9
             Manual          376
Iki 4000     Automatic        13
             Manual          408
Iki 5000     Automatic        13
             Manual          263
Iki 7000     Automatic        16
             Manual          368
Nuo 20000    Automatic        37
             Manual            5
Name: price_EUR, dtype: int64

Hipotezė 2

Individualūs savininkai automobilius parduoda pigiau (mažiau reklamos, nesuteikia garantijos, sensni, patikimumas ir kt.).

(Visiškai pasitvirtino)

car.head(1)

	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make	price_EUR	price_group	km_group
0	2007	60000	70000	Petrol	Individual	Manual	First Owner	Maruti	720.0	Iki 1000	Iki 100000 km

print(car.groupby(['seller_type'])['price_EUR'].mean().round(0))

seller_type
Dealer               8662.0
Individual           5094.0
Trustmark Dealer    10979.0
Name: price_EUR, dtype: float64

Hipotezė 3

Norint įsigyti automobilį nuvažiavusį ne daugiau 50 K km ir ne senesnį kaip 3 metų amžiaus, dažniausiai reikia "pasiruošti" bent 10 k EUR.

(Dažniausiai taip)

car.head(1)

	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make	price_EUR	price_group	km_group
0	2007	60000	70000	Petrol	Individual	Manual	First Owner	Maruti	720.0	Iki 1000	Iki 100000 km

car[car['km_driven'] <= 50000].groupby(['year'])['price_EUR'].mean().round(0)

year
1996     3000.0
1998     5112.0
1999     1100.0
2000      891.0
2001     2314.0
2002     1245.0
2003      679.0
2004     1410.0
2005     1107.0
2006     4105.0
2007     1832.0
2008     2108.0
2009     2953.0
2010     3446.0
2011     3023.0
2012     4097.0
2013     5483.0
2014     5963.0
2015     6242.0
2016     6534.0
2017     9195.0
2018    10928.0
2019    13021.0
2020    10043.0
Name: price_EUR, dtype: float64

car.plot.bar(x='price_group', y='km_driven', title='Hipotezė 3')

<AxesSubplot:title={'center':'Hipotezė 3'}, xlabel='price_group'>

Hipotezė 4

Jei pirkėjas turi iki 3 k EUR, reikėtų tikėtis automobilio su "mechanine" pavarų dėže ir nuvažiavusio bent 100 k km.

(Visiškai nepasitvirtino)

car.head(1)

	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make	price_EUR	price_group	km_group
0	2007	60000	70000	Petrol	Individual	Manual	First Owner	Maruti	720.0	Iki 1000	Iki 100000 km

def if_km(x):
    if x <= 100000:
        return "Iki 100000 km"
    else:
        return "nuo 100000 km"

reiksmes_vektorizuoti = np.vectorize(if_km)
sugrupuoti = reiksmes_vektorizuoti(car['km_driven'])

car['km_group'] = sugrupuoti
car.head(6)

	year	selling_price	km_driven	fuel	seller_type	transmission	owner	make	price_EUR	price_group	km_group
0	2007	60000	70000	Petrol	Individual	Manual	First Owner	Maruti	720.0	Iki 1000	Iki 100000 km
1	2007	135000	50000	Petrol	Individual	Manual	First Owner	Maruti	1620.0	Iki 2000	Iki 100000 km
2	2012	600000	100000	Diesel	Individual	Manual	First Owner	Hyundai	7200.0	Iki 10000	Iki 100000 km
3	2017	250000	46000	Petrol	Individual	Manual	First Owner	Datsun	3000.0	Iki 4000	Iki 100000 km
4	2014	450000	141000	Diesel	Individual	Manual	Second Owner	Honda	5400.0	Iki 7000	nuo 100000 km
5	2007	140000	125000	Petrol	Individual	Manual	First Owner	Maruti	1680.0	Iki 2000	nuo 100000 km

car[(car['price_EUR'] < 3000)].groupby(['km_group', 'transmission'])['price_EUR'].mean().round(0)

km_group       transmission
Iki 100000 km  Automatic       2020.0
               Manual          1786.0
nuo 100000 km  Automatic       1824.0
               Manual          1691.0
Name: price_EUR, dtype: float64

Prognozė

Vertinant automobilio ridą, kokia turėtų būti jo kaina.

car.plot.scatter(x='km_driven', y='price_EUR')

<AxesSubplot:xlabel='km_driven', ylabel='price_EUR'>

sns.jointplot(x='km_driven', y='price_EUR', data=car)
sns.displot(car['price_EUR'])

<seaborn.axisgrid.FacetGrid at 0x1b1f3b51490>

x = car[['km_driven']]
y = car['price_EUR']
X = x

model = LinearRegression()
model.fit(x, y)

LinearRegression()

prediction = model.predict(x)
prediction

array([5941.22175259, 6513.63305302, 5082.60480194, ..., 5569.15440731,
       5368.81045215, 6799.83870324])

model.predict([[10000]]).round(0)

array([7658.])

r_squared = model.score(x, y)
r_squared

0.036974943094581936

plt.scatter(x, y)
plt.plot(x, prediction, 'green')

[<matplotlib.lines.Line2D at 0x1b1f3962af0>]

car.to_csv('car_deho_india_original_for_Tableau_cool.csv')