Resale Price Prediction

# import the libraries
import re
import scipy
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

from sklearn.pipeline import make_pipeline
from sklearn.decomposition import TruncatedSVD
from sklearn.model_selection import train_test_split
from scipy.sparse import coo_matrix, hstack
from sklearn.preprocessing import StandardScaler, MinMaxScaler
from sklearn.preprocessing import OneHotEncoder
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.feature_extraction.text import CountVectorizer

from keras import backend as K
import tensorflow as tf
from tensorflow import keras
from keras.layers import Dense, Input, Dropout
from keras.models import Model

from utils import *

import warnings
warnings.filterwarnings("ignore")

plt.style.use('fivethirtyeight')
plt.style.use('seaborn-notebook')

%config InlineBackend.figure_format = 'retina'
%reload_ext autoreload
%autoreload 2

df = pd.read_pickle('./data/df_cleaned.p')

colname_map = {'PRC':'BRAND', 'PARTNO':'PARTNO','UNIT RESALE':'UNITRESALE',
               'ORIG ORDER QTY':'ORDERQTY', 'NEW UNIT COST':'UNITCOST'}
df = prepare_data(df, colname_map)

df.head()

	BRAND	PARTNO	QUANTITY	UNITRESALE	UNITCOST
0	2	53001176-1 REV A	25-49	209.66	107.65
1	2	53001176-1 REV A	25-49	209.66	99.51
2	2	61-82477-8	25-49	76.75	60
3	2	AA1208K08	25-49	66.12	50.8
4	2	AA1208K08X	50-99	66.21	52

df, fitted_lambda = scale_price(df)

df.head()

	BRAND	PARTNO	QUANTITY	UNITCOST	RESALE
0	2	61-82477-8	25-49	60	3.850272
1	2	AA1208K08	25-49	50.8	3.733058
2	2	AA1208K08X	50-99	52	3.734132
3	2	AA67006-4KA	50-99	13.9	2.559783
4	2	AA67006-4KA	25-49	13.5	2.686291

CV1 = CountVectorizer(stop_words=None, 
                      max_df=1.0, 
                      min_df=100, 
                      ngram_range=(1,1),
                      binary=True, 
                      analyzer='char')

CV1.fit(list(set(df['PARTNO'].tolist())))
X1 = CV1.transform(df['PARTNO'].tolist())
X1

<432960x45 sparse matrix of type '<class 'numpy.int64'>'
    with 3493797 stored elements in Compressed Sparse Row format>

# CV1.vocabulary_

CV2 = CountVectorizer(stop_words=None, 
                      max_df=0.8, 
                      min_df=100, 
                      ngram_range=(2,6), 
                      binary=True,
                      analyzer='char')
CV2.fit(list(set(df['PARTNO'].tolist())))
X2 = CV2.transform(df['PARTNO'].tolist())
X2

<432960x5430 sparse matrix of type '<class 'numpy.int64'>'
    with 9427277 stored elements in Compressed Sparse Row format>

def tokenizer(text):
  text = text.lower()
  rx1 = r"(?i)(?:(?<=\d)(?=[a-z])|(?<=[a-z])(?=\d))"
  text = re.sub(rx1,' ', text)
  text = re.sub(r'[^a-z0-9]',' ', text)
  text = ' '.join(text.split())
  text = text.split()
  return text

CV3 = TfidfVectorizer(stop_words=None, 
                      max_df=0.5, 
                      min_df=100, 
                      ngram_range=(1,5), 
                      binary=False,
                      analyzer='word',
                      tokenizer=tokenizer)
CV3.fit(list(set(df['PARTNO'].tolist())))
X3 = CV3.transform(df['PARTNO'].tolist())
X3

<432960x1007 sparse matrix of type '<class 'numpy.float64'>'
    with 1715717 stored elements in Compressed Sparse Row format>

enc = OneHotEncoder()
ohecols = ['BRAND','QUANTITY']
enc.fit(df[ohecols])
X4 = enc.transform(df[ohecols])
X4

<432960x577 sparse matrix of type '<class 'numpy.float64'>'
    with 865920 stored elements in Compressed Sparse Row format>

X = hstack([X1, X2, X3, X4])
X

<432960x7059 sparse matrix of type '<class 'numpy.float64'>'
    with 15502711 stored elements in COOrdinate format>

Y = df['RESALE'].values
Y = Y.reshape(-1,1)

X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.1, random_state=40)
print("Training Records {}, Testing Records: {}".format(X_train.shape[0],
                                                        X_test.shape[0]))

Training Records 389664, Testing Records: 43296

batch_size = 2048
epochs = 75

inputs = Input(shape=(X_train.shape[1],), sparse=True)
L = Dense(512, activation='relu')(inputs)
L = Dropout(0.5)(L)
L = Dense(10, activation='relu')(L)
outputs = Dense(y_train.shape[1])(L)
model = Model(inputs=inputs, outputs=outputs)
model.compile(loss='mse', optimizer='adam', metrics=['mae'])
model.summary()

Model: "functional_3"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_2 (InputLayer)         [(None, 7059)]            0         
_________________________________________________________________
dense_3 (Dense)              (None, 512)               3614720   
_________________________________________________________________
dropout_1 (Dropout)          (None, 512)               0         
_________________________________________________________________
dense_4 (Dense)              (None, 10)                5130      
_________________________________________________________________
dense_5 (Dense)              (None, 1)                 11        
=================================================================
Total params: 3,619,861
Trainable params: 3,619,861
Non-trainable params: 0
_________________________________________________________________

history = model.fit(nn_batch_generator(X_train, y_train, batch_size),
          steps_per_epoch=len(y_train)//batch_size, 
          validation_data=nn_batch_generator(X_test, y_test, batch_size),
          validation_steps=len(y_test)//batch_size, 
          epochs=100,
          workers=-1, 
          use_multiprocessing=True)

Epoch 1/100
190/190 [==============================] - 22s 114ms/step - loss: 0.8163 - mae: 0.6779 - val_loss: 0.4789 - val_mae: 0.5142
Epoch 2/100
190/190 [==============================] - 21s 111ms/step - loss: 0.4670 - mae: 0.5137 - val_loss: 0.4088 - val_mae: 0.4675
Epoch 3/100
190/190 [==============================] - 21s 111ms/step - loss: 0.3991 - mae: 0.4744 - val_loss: 0.3682 - val_mae: 0.4406
Epoch 4/100
190/190 [==============================] - 21s 112ms/step - loss: 0.3548 - mae: 0.4472 - val_loss: 0.3426 - val_mae: 0.4222
Epoch 5/100
190/190 [==============================] - 21s 111ms/step - loss: 0.3244 - mae: 0.4278 - val_loss: 0.3288 - val_mae: 0.4123
Epoch 6/100
190/190 [==============================] - 21s 112ms/step - loss: 0.3011 - mae: 0.4124 - val_loss: 0.3146 - val_mae: 0.4022
Epoch 7/100
190/190 [==============================] - 22s 115ms/step - loss: 0.2838 - mae: 0.4004 - val_loss: 0.3066 - val_mae: 0.3963
Epoch 8/100
190/190 [==============================] - 21s 112ms/step - loss: 0.2701 - mae: 0.3905 - val_loss: 0.3014 - val_mae: 0.3924
Epoch 9/100
190/190 [==============================] - 21s 111ms/step - loss: 0.2593 - mae: 0.3820 - val_loss: 0.2951 - val_mae: 0.3861
Epoch 10/100
190/190 [==============================] - 22s 115ms/step - loss: 0.2493 - mae: 0.3746 - val_loss: 0.2937 - val_mae: 0.3852
Epoch 11/100
190/190 [==============================] - 21s 112ms/step - loss: 0.2419 - mae: 0.3688 - val_loss: 0.2899 - val_mae: 0.3798
Epoch 12/100
190/190 [==============================] - 21s 113ms/step - loss: 0.2353 - mae: 0.3633 - val_loss: 0.2867 - val_mae: 0.3771
Epoch 13/100
190/190 [==============================] - 22s 116ms/step - loss: 0.2279 - mae: 0.3574 - val_loss: 0.2877 - val_mae: 0.3767
Epoch 14/100
190/190 [==============================] - 21s 112ms/step - loss: 0.2238 - mae: 0.3538 - val_loss: 0.2823 - val_mae: 0.3730
Epoch 15/100
190/190 [==============================] - 21s 112ms/step - loss: 0.2196 - mae: 0.3498 - val_loss: 0.2821 - val_mae: 0.3720
Epoch 16/100
190/190 [==============================] - 22s 114ms/step - loss: 0.2142 - mae: 0.3460 - val_loss: 0.2788 - val_mae: 0.3696
Epoch 17/100
190/190 [==============================] - 21s 111ms/step - loss: 0.2105 - mae: 0.3426 - val_loss: 0.2801 - val_mae: 0.3698
Epoch 18/100
190/190 [==============================] - 21s 112ms/step - loss: 0.2072 - mae: 0.3397 - val_loss: 0.2770 - val_mae: 0.3669
Epoch 19/100
190/190 [==============================] - 22s 115ms/step - loss: 0.2039 - mae: 0.3369 - val_loss: 0.2790 - val_mae: 0.3690
Epoch 20/100
190/190 [==============================] - 21s 112ms/step - loss: 0.2014 - mae: 0.3345 - val_loss: 0.2755 - val_mae: 0.3657
Epoch 21/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1980 - mae: 0.3314 - val_loss: 0.2754 - val_mae: 0.3660
Epoch 22/100
190/190 [==============================] - 21s 108ms/step - loss: 0.1952 - mae: 0.3290 - val_loss: 0.2757 - val_mae: 0.3649
Epoch 23/100
190/190 [==============================] - 22s 114ms/step - loss: 0.1933 - mae: 0.3275 - val_loss: 0.2731 - val_mae: 0.3633
Epoch 24/100
190/190 [==============================] - 21s 111ms/step - loss: 0.1913 - mae: 0.3254 - val_loss: 0.2734 - val_mae: 0.3632
Epoch 25/100
190/190 [==============================] - 20s 104ms/step - loss: 0.1880 - mae: 0.3225 - val_loss: 0.2721 - val_mae: 0.3617
Epoch 26/100
190/190 [==============================] - 21s 111ms/step - loss: 0.1870 - mae: 0.3213 - val_loss: 0.2732 - val_mae: 0.3626
Epoch 27/100
190/190 [==============================] - 22s 114ms/step - loss: 0.1838 - mae: 0.3186 - val_loss: 0.2698 - val_mae: 0.3588
Epoch 28/100
190/190 [==============================] - 21s 113ms/step - loss: 0.1823 - mae: 0.3173 - val_loss: 0.2708 - val_mae: 0.3589
Epoch 29/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1808 - mae: 0.3155 - val_loss: 0.2712 - val_mae: 0.3599
Epoch 30/100
190/190 [==============================] - 19s 102ms/step - loss: 0.1790 - mae: 0.3140 - val_loss: 0.2727 - val_mae: 0.3607
Epoch 31/100
190/190 [==============================] - 21s 111ms/step - loss: 0.1772 - mae: 0.3126 - val_loss: 0.2698 - val_mae: 0.3588
Epoch 32/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1756 - mae: 0.3108 - val_loss: 0.2703 - val_mae: 0.3584
Epoch 33/100
190/190 [==============================] - 19s 99ms/step - loss: 0.1747 - mae: 0.3099 - val_loss: 0.2693 - val_mae: 0.3575
Epoch 34/100
190/190 [==============================] - 21s 111ms/step - loss: 0.1733 - mae: 0.3085 - val_loss: 0.2703 - val_mae: 0.3578
Epoch 35/100
190/190 [==============================] - 21s 108ms/step - loss: 0.1724 - mae: 0.3077 - val_loss: 0.2704 - val_mae: 0.3583
Epoch 36/100
190/190 [==============================] - 19s 99ms/step - loss: 0.1706 - mae: 0.3063 - val_loss: 0.2694 - val_mae: 0.3580
Epoch 37/100
190/190 [==============================] - 21s 111ms/step - loss: 0.1689 - mae: 0.3045 - val_loss: 0.2687 - val_mae: 0.3573
Epoch 38/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1678 - mae: 0.3033 - val_loss: 0.2690 - val_mae: 0.3572
Epoch 39/100
190/190 [==============================] - 19s 98ms/step - loss: 0.1668 - mae: 0.3023 - val_loss: 0.2698 - val_mae: 0.3578
Epoch 40/100
190/190 [==============================] - 21s 110ms/step - loss: 0.1655 - mae: 0.3008 - val_loss: 0.2687 - val_mae: 0.3556
Epoch 41/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1643 - mae: 0.3001 - val_loss: 0.2702 - val_mae: 0.3569
Epoch 42/100
190/190 [==============================] - 19s 98ms/step - loss: 0.1637 - mae: 0.2992 - val_loss: 0.2698 - val_mae: 0.3567
Epoch 43/100
190/190 [==============================] - 21s 111ms/step - loss: 0.1626 - mae: 0.2980 - val_loss: 0.2703 - val_mae: 0.3570
Epoch 44/100
190/190 [==============================] - 21s 113ms/step - loss: 0.1616 - mae: 0.2973 - val_loss: 0.2695 - val_mae: 0.3552
Epoch 45/100
190/190 [==============================] - 20s 104ms/step - loss: 0.1610 - mae: 0.2965 - val_loss: 0.2666 - val_mae: 0.3532
Epoch 46/100
190/190 [==============================] - 19s 101ms/step - loss: 0.1594 - mae: 0.2948 - val_loss: 0.2683 - val_mae: 0.3530
Epoch 47/100
190/190 [==============================] - 19s 102ms/step - loss: 0.1587 - mae: 0.2939 - val_loss: 0.2688 - val_mae: 0.3532
Epoch 48/100
190/190 [==============================] - 21s 110ms/step - loss: 0.1576 - mae: 0.2931 - val_loss: 0.2676 - val_mae: 0.3528
Epoch 49/100
190/190 [==============================] - 21s 112ms/step - loss: 0.1565 - mae: 0.2918 - val_loss: 0.2679 - val_mae: 0.3528
Epoch 50/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1558 - mae: 0.2911 - val_loss: 0.2676 - val_mae: 0.3522
Epoch 51/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1549 - mae: 0.2900 - val_loss: 0.2684 - val_mae: 0.3535
Epoch 52/100
190/190 [==============================] - 21s 108ms/step - loss: 0.1543 - mae: 0.2895 - val_loss: 0.2707 - val_mae: 0.3553
Epoch 53/100
190/190 [==============================] - 21s 110ms/step - loss: 0.1535 - mae: 0.2886 - val_loss: 0.2673 - val_mae: 0.3528
Epoch 54/100
190/190 [==============================] - 20s 106ms/step - loss: 0.1531 - mae: 0.2880 - val_loss: 0.2672 - val_mae: 0.3523
Epoch 55/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1519 - mae: 0.2870 - val_loss: 0.2673 - val_mae: 0.3528
Epoch 56/100
190/190 [==============================] - 21s 109ms/step - loss: 0.1513 - mae: 0.2862 - val_loss: 0.2670 - val_mae: 0.3528
Epoch 57/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1503 - mae: 0.2852 - val_loss: 0.2684 - val_mae: 0.3535
Epoch 58/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1497 - mae: 0.2847 - val_loss: 0.2678 - val_mae: 0.3526
Epoch 59/100
190/190 [==============================] - 21s 110ms/step - loss: 0.1489 - mae: 0.2837 - val_loss: 0.2668 - val_mae: 0.3514
Epoch 60/100
190/190 [==============================] - 20s 106ms/step - loss: 0.1480 - mae: 0.2828 - val_loss: 0.2681 - val_mae: 0.3531
Epoch 61/100
190/190 [==============================] - 21s 108ms/step - loss: 0.1476 - mae: 0.2822 - val_loss: 0.2675 - val_mae: 0.3520
Epoch 62/100
190/190 [==============================] - 21s 110ms/step - loss: 0.1474 - mae: 0.2818 - val_loss: 0.2674 - val_mae: 0.3518
Epoch 63/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1465 - mae: 0.2809 - val_loss: 0.2670 - val_mae: 0.3508
Epoch 64/100
190/190 [==============================] - 21s 108ms/step - loss: 0.1462 - mae: 0.2808 - val_loss: 0.2655 - val_mae: 0.3499
Epoch 65/100
190/190 [==============================] - 21s 110ms/step - loss: 0.1455 - mae: 0.2800 - val_loss: 0.2678 - val_mae: 0.3510
Epoch 66/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1445 - mae: 0.2789 - val_loss: 0.2679 - val_mae: 0.3507
Epoch 67/100
190/190 [==============================] - 20s 103ms/step - loss: 0.1438 - mae: 0.2780 - val_loss: 0.2680 - val_mae: 0.3507
Epoch 68/100
190/190 [==============================] - 20s 106ms/step - loss: 0.1434 - mae: 0.2777 - val_loss: 0.2683 - val_mae: 0.3514
Epoch 69/100
190/190 [==============================] - 22s 113ms/step - loss: 0.1424 - mae: 0.2769 - val_loss: 0.2693 - val_mae: 0.3516
Epoch 70/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1426 - mae: 0.2768 - val_loss: 0.2679 - val_mae: 0.3508
Epoch 71/100
190/190 [==============================] - 20s 105ms/step - loss: 0.1410 - mae: 0.2752 - val_loss: 0.2672 - val_mae: 0.3505
Epoch 72/100
190/190 [==============================] - 21s 113ms/step - loss: 0.1410 - mae: 0.2749 - val_loss: 0.2663 - val_mae: 0.3493
Epoch 73/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1404 - mae: 0.2742 - val_loss: 0.2681 - val_mae: 0.3512
Epoch 74/100
190/190 [==============================] - 20s 105ms/step - loss: 0.1398 - mae: 0.2738 - val_loss: 0.2666 - val_mae: 0.3490
Epoch 75/100
190/190 [==============================] - 22s 114ms/step - loss: 0.1400 - mae: 0.2736 - val_loss: 0.2663 - val_mae: 0.3497
Epoch 76/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1393 - mae: 0.2731 - val_loss: 0.2665 - val_mae: 0.3494
Epoch 77/100
190/190 [==============================] - 20s 105ms/step - loss: 0.1388 - mae: 0.2724 - val_loss: 0.2672 - val_mae: 0.3505
Epoch 78/100
190/190 [==============================] - 22s 114ms/step - loss: 0.1386 - mae: 0.2720 - val_loss: 0.2678 - val_mae: 0.3499
Epoch 79/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1382 - mae: 0.2717 - val_loss: 0.2678 - val_mae: 0.3497
Epoch 80/100
190/190 [==============================] - 20s 104ms/step - loss: 0.1377 - mae: 0.2711 - val_loss: 0.2675 - val_mae: 0.3497
Epoch 81/100
190/190 [==============================] - 22s 113ms/step - loss: 0.1374 - mae: 0.2709 - val_loss: 0.2669 - val_mae: 0.3495
Epoch 82/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1370 - mae: 0.2702 - val_loss: 0.2674 - val_mae: 0.3496
Epoch 83/100
190/190 [==============================] - 20s 104ms/step - loss: 0.1362 - mae: 0.2692 - val_loss: 0.2675 - val_mae: 0.3504
Epoch 84/100
190/190 [==============================] - 21s 113ms/step - loss: 0.1358 - mae: 0.2690 - val_loss: 0.2664 - val_mae: 0.3496
Epoch 85/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1351 - mae: 0.2681 - val_loss: 0.2692 - val_mae: 0.3511
Epoch 86/100
190/190 [==============================] - 20s 104ms/step - loss: 0.1352 - mae: 0.2679 - val_loss: 0.2671 - val_mae: 0.3498
Epoch 87/100
190/190 [==============================] - 21s 113ms/step - loss: 0.1346 - mae: 0.2675 - val_loss: 0.2687 - val_mae: 0.3503
Epoch 88/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1343 - mae: 0.2672 - val_loss: 0.2691 - val_mae: 0.3494
Epoch 89/100
190/190 [==============================] - 20s 103ms/step - loss: 0.1337 - mae: 0.2667 - val_loss: 0.2687 - val_mae: 0.3503
Epoch 90/100
190/190 [==============================] - 20s 104ms/step - loss: 0.1335 - mae: 0.2660 - val_loss: 0.2689 - val_mae: 0.3502
Epoch 91/100
190/190 [==============================] - 20s 106ms/step - loss: 0.1334 - mae: 0.2659 - val_loss: 0.2686 - val_mae: 0.3494
Epoch 92/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1331 - mae: 0.2657 - val_loss: 0.2674 - val_mae: 0.3489
Epoch 93/100
190/190 [==============================] - 20s 104ms/step - loss: 0.1326 - mae: 0.2649 - val_loss: 0.2664 - val_mae: 0.3486
Epoch 94/100
190/190 [==============================] - 20s 105ms/step - loss: 0.1320 - mae: 0.2644 - val_loss: 0.2668 - val_mae: 0.3486
Epoch 95/100
190/190 [==============================] - 20s 107ms/step - loss: 0.1322 - mae: 0.2645 - val_loss: 0.2683 - val_mae: 0.3494
Epoch 96/100
190/190 [==============================] - 20s 104ms/step - loss: 0.1313 - mae: 0.2637 - val_loss: 0.2675 - val_mae: 0.3496
Epoch 97/100
190/190 [==============================] - 20s 106ms/step - loss: 0.1313 - mae: 0.2636 - val_loss: 0.2666 - val_mae: 0.3485
Epoch 98/100
190/190 [==============================] - 20s 108ms/step - loss: 0.1309 - mae: 0.2633 - val_loss: 0.2679 - val_mae: 0.3493
Epoch 99/100
190/190 [==============================] - 20s 103ms/step - loss: 0.1309 - mae: 0.2629 - val_loss: 0.2670 - val_mae: 0.3487
Epoch 100/100
190/190 [==============================] - 20s 106ms/step - loss: 0.1301 - mae: 0.2622 - val_loss: 0.2667 - val_mae: 0.3485

model.save('./models/model_201203.h5')

hist_df = pd.DataFrame(history.history) 
hist_csv_file = './outputs/history.csv'
with open(hist_csv_file, mode='w') as f:
  hist_df.to_csv(f)

from scipy.special import inv_boxcox
from sklearn.metrics import r2_score, median_absolute_error, mean_absolute_error

y_pred = model.predict(X_test).flatten()
a = inv_boxcox(y_test.flatten(), fitted_lambda)
b = inv_boxcox(y_pred.flatten(), fitted_lambda)
print('r2_score: ', r2_score(a, b))
print('median_absolute_error: ', median_absolute_error(a, b))
print('mean_absolute_error', mean_absolute_error(a, b))
out2 = pd.DataFrame({'y_true':inv_boxcox(y_test.flatten(), fitted_lambda), 'y_pred':inv_boxcox(y_pred.flatten(), fitted_lambda)})

r2_score:  0.7725025811056968
median_absolute_error:  0.49407594919204767
mean_absolute_error 3.357431710265042

out2.head()

	y_true	y_pred
0	4.65	10.811751
1	7.6	7.666917
2	1.1	0.746361
3	0.72	0.291657
4	41.4	31.236202

_, out1 = train_test_split(df, test_size=0.1, random_state=40)
out1['RESALE'] = out2.y_true.values
out1['PRED'] = out2.y_pred.values
out1.to_csv('./outputs/result.csv', index=False)

out1.sample(10)

	BRAND	PARTNO	QUANTITY	UNITCOST	RESALE	PRED
174877	155	TLM-6X1C-12	2500-4999	0.09	0.15	0.170021
39616	59	MS3498-9	500-999	1.12	1.5	1.150005
18036	30	8106-A-0440-17	500-999	0.17	0.4	0.61341
406828	662	250-0201-01	50-99	3.01	5	4.733656
78736	78	SC-16-SB	05-Sep	59.05	84.35	91.922928
116973	116	NAS6206-18	01-Apr	2.18	25	24.017282
59365	63	MS21087-4	25-49	3.5	5.22	7.015954
256983	350	0326010.HXP	Oct-24	0.4	4.5	4.300601
211941	212	MS91528-1F2B	25-49	2.55	4.4	4.558315
334493	525	M24243/1A404	5000-9999	0.12	0.26	0.203509