Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

HouseSales Quarterly Solutions

Download as xlsx, pdf, or txt
Download as xlsx, pdf, or txt
You are on page 1of 11

Year Quarter Period # Sold Data thanks to:

2015 1 1 513237 https://www.redfin.com/blog/data-cen


2015 2 2 739297
2015 3 3 753122
Total Number of Houses sold in the United States p
2015 4 4 605205
(from 2015 to present)
2016 1 5 525558
2016 2 6 802271 900000

2016 3 7 761358 850000

2016 4 8 671747 800000

2017 1 9 572641 750000

2017 2 10 836834 700000

2017 3 11 784570 650000

2017 4 12 684706 600000


2018 1 13 578054 550000
2018 2 14 814794 500000
2018 3 15 794354 450000
2018 4 16 648794 400000
0 5 10 15
2019 1 17 560664
2019 2 18 828881
2019 3 19 795389
2019 4 20 717534
2020 1 21 600812
ww.redfin.com/blog/data-center/

ld in the United States per Quarter


015 to present)

10 15 20
Period Year Quarter # Sold Level Trend Seasonal Forecast Error
1 2015 1 513237
2 2015 2 739297
3 2015 3 753122
4 2015 4 605205
5 2016 1 525558
6 2016 2 802271
7 2016 3 761358
8 2016 4 671747
9 2017 1 572641
10 2017 2 836834
11 2017 3 784570
12 2017 4 684706
13 2018 1 578054
14 2018 2 814794
15 2018 3 794354
16 2018 4 648794
17 2019 1 560664
18 2019 2 828881
19 2019 3 795389
20 2019 4 717534
21 2020 1 600812
22 2020 2
23 2020 3
24 2020 4
alpha 0.1
beta 0.1
gamma 0.1
RMSE > 18350.15 <--- Regression is a better forecasting method on this data
Year Quarter # Sold Coded Time Quarter 2 Quarter 3 Quarter 4 Forecasts Errors
2015 1 513237 0 0 0 0 517166 -3929
2015 2 739297 1 1 0 0 771353 -32056
2015 3 753122 2 0 1 0 744696 8426
2015 4 605205 3 0 0 1 632535 -27330
2016 1 525558 4 0 0 0 533698 -8140
2016 2 802271 5 1 0 0 787884 14387
2016 3 761358 6 0 1 0 761227 131
2016 4 671747 7 0 0 1 649066 22681
2017 1 572641 8 0 0 0 550229 22412
2017 2 836834 9 1 0 0 804415 32419
2017 3 784570 10 0 1 0 777759 6811
2017 4 684706 11 0 0 1 665597 19109
2018 1 578054 12 0 0 0 566760 11294
2018 2 814794 13 1 0 0 820947 -6153
2018 3 794354 14 0 1 0 794290 64
2018 4 648794 15 0 0 1 682128 -33334
2019 1 560664 16 0 0 0 583291 -22627
2019 2 828881 17 1 0 0 837478 -8597
2019 3 795389 18 0 1 0 810821 -15432
2019 4 717534 19 0 0 1 698659 18875
2020 1 600812 20 0 0 0 599822 990
2020 2 21 1 0 0 854009
2020 3 22 0 1 0 827352
2020 4 23 0 0 1 715191
2021 1 24 0 0 0 616353
2021 2 25 1 0 0 870540
2021 3 26 0 1 0 843883
2021 4 27 0 0 1 731722
Data thanks to:
https://www.redfin.com/blog/data-center/ Total Number of Houses sold in the United States pe
Quarter (from 2015 to present)
RMSE = 18350.15 < ??? 900000
850000
Regression is better on this data set 800000
than Holt Winter's Method 750000
Coefficients 700000
650000
Intercept 517166.5 b0 600000
Coded Time 4132.784 b1 550000
500000
Quarter 2 250053.9 b2
450000
Quarter 3 219264.3 b3 400000
-3 2 7 12 17
Quarter 4 102970.1 b4

1. What is the forecast # houses sold for Quarter 1 of 2015? 517166.5


2. On average, how much do the house sales increase each quarter
(holding seasonal effects constant)? 4132.784
3. How many more houses are sold, on average, in Quarter 2 than
Quarter 1 (holding all other variables constant)? 250053.9
sold in the United States per
2015 to present)

12 17 22

7166.5
arter

han
SUMMARY OUTPUT

Regression Statistics Coded Ti


Multiple R 0.984527 40000
R Square 0.969293
20000

Residuals
Adjusted R 0.961616
Standard E 21022.74 0
0
Observatio 21 -20000
-40000
ANOVA
df SS MS F Significance F
Regression 4 2.23E+11 5.58E+10 126.2618 6.921E-12
Residual 16 7.07E+09 4.42E+08
Total 20 2.3E+11

Coefficients
Standard Error t Stat P-value Lower 95%Upper 95%Lower 95.0%
Upper 95.0%
Intercept 517166.5 11480.91 45.04577 2.766E-18 492828 541504.9 492828 541504.9
Coded Tim 4132.784 762.575 5.419512 5.674E-05 2516.197 5749.371 2516.197 5749.371
Quarter 2 250053.9 12752.72 19.60788 1.298E-12 223019.3 277088.4 223019.3 277088.4
Quarter 3 219264.3 12729.9 17.22435 9.449E-12 192278.1 246250.5 192278.1 246250.5
Quarter 4 102970.1 12752.72 8.074361 4.917E-07 75935.52 130004.6 75935.52 130004.6

RESIDUAL OUTPUT PROBABILITY OUTPUT

Observation
Predicted # Sold
Residuals Percentile # Sold
1 517166.5 -3929.491 2.380952 513237
2 771353.1 -32056.13 7.142857 525558
3 744696.3 8425.674 11.90476 560664
4 632534.9 -27329.93 16.66667 572641
5 533697.6 -8139.628 21.42857 578054
6 787884.3 14386.74 26.19048 600812
7 761227.5 130.5368 30.95238 605205
8 649066.1 22680.94 35.71429 648794
9 550228.8 22412.24 40.47619 671747
10 804415.4 32418.6 45.2381 684706
11 777758.6 6811.4 50 717534
12 665597.2 19108.8 54.7619 739297
13 566759.9 11294.1 59.52381 753122
14 820946.5 -6152.537 64.28571 761358
15 794289.7 64.26316 69.04762 784570
16 682128.3 -33334.34 73.80952 794354
17 583291 -22627 78.57143 795389
18 837477.7 -8596.674 83.33333 802271
19 810820.9 -15431.87 88.09524 814794
20 698659.5 18874.53 92.85714 828881
21 599822.2 989.8246 97.61905 836834
Coded Time Residual Plot
40000
20000
Quarter 2 Residual Plot
Residuals

40000
0
0 200005 10
Quarter
15
3 20
Residual
25
Plot
-20000
Residuals

40000
-40000
0
0 0.2
20000 0.4
Quarter 4 Residual Plot
0.6 0.8 1 1.2
-20000
Residuals

Coded Time40000
-40000
0
0 20000 0.2 0.4
Coded Time Line Fit Plot
0.6 0.8 1 1.2
-20000
Residuals

Quarter 2 1000000
-40000
0
0
800000
0.2 0.4
Quarter 2 Line Fit Plot
0.6 0.8 1 1.2
-20000 600000
Quarter 3 1000000 # Sold
# Sold

-40000
400000
200000
800000 Quarter 3 Line Fit Plot
Predicted # Sold
600000
Quarter 4 1000000 # Sold
0
# Sold

0
400000
5
200000
10 800000
15 20 25 Quarter 4 Line Fit Plot
Predicted # Sold
Coded0 Time600000 1000000 # Sold
# Sold 400000 Normal Probability P
Predicted # Sold
0 0.2 0.4 0.6800000
0.8 1 1.2
200000
Quarter 600000 # Sold
0 2

# Sold
1000000
400000 Predicted
0 0.2 0.4 0.6800000
0.8 1 1.2
200000
Quarter 600000
0 3

# Sold
400000
0 0.2 0.4 0.6 0.8 1 1.2
200000
Quarter
0 4
0 20 40 60 8
Sample Percentile
ot
Fit Plot
# Sold
4 Line Fit Plot
Predicted # Sold
# Sold
ormal Probability Plot
Predicted # Sold
# Sold
1 1.2 Predicted # Sold

0.6 0.8 1 1.2


er 4
0 20 40 60 80 100 120
Sample Percentile

You might also like