Power Law:(36)の続き。Continuation of Power Law:(36).
http://humanbeing-etcman.blogspot.com/2008/12/power-law36rlinear-approximation2.html
R^2により、べき乗傾向の範囲の方針を決定したい。
I want to decide the policy within the range of the Power Law tendency by R^2.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~(2008/12/21,06:45-)
[1]単純に、R^2の最大値を求める。Simply, the maximum value of R^2 is requested.
：Power Law:(36)のデータ：kinji_ALL を使う。The data of Power Law:(36): Kinji_ALL is used.
---
[1-1]先頭２件がNA値なので、3番から、最後(この場合、69)までを対象にする。
Because two heads are the NA values, everything from the 3rd to the end (69 in this case) is targeted.
===
参考）
http://www.is.titech.ac.jp/~shimo/class/doc/r-tips.pdf
：pdf内をキーワードで検索する。
：参考ページ）p.34, p.54, p.95, p.143
===
> max(kinji_ALL$KINJI_R2[3:69])
[1] 0.984492
>
---
[1-2]最大値は、配列の何番目か？Where of the array is the maximum value?
===
参考）
http://www.okada.jp.org/RWiki/?%B9%D4%CE%F3Tips%C2%E7%C1%B4#content_1_58
===
> which.max(kinji_ALL$KINJI_R2[3:69])
[1] 26
>
：先頭のNA値２件をスキップしているので、実際の順番は＋２となる。
:The actual order becomes +2 because it skips the first two NA values.
> which.max(kinji_ALL$KINJI_R2[3:69]) + 2
[1] 28
>
---
[1-3]最大値を求める最終形は以下。The final type from which the maximum value is requested is the following.
> max_order = which.max(kinji_ALL$KINJI_R2[3:69]) + 2
> max_r2 = kinji_ALL$KINJI_R2[max_order]
> max_order
[1] 28
> max_r2
[1] 0.984492
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~(2008/12/19,14:44-)
[2]２８位は、本：「業界地図」との絡みはあるか？ Does 28th place have twining with the book "Industry map"?
「会社四季報」業界地図〈2009年版〉 東洋経済新報社 東経= (単行本 - 2008/8)
：20081212、東証一部の食品、終値の時価総額に関する順位
[2-1]上位３社, Top 3
2914,20081212,ＪＴ,311000,10000,3110000000,1,0,21.85788856
2503,20081212,キリン,1082,984508,1065237656,2,0.693147181,20.78646376
2502,20081212,アサヒ,1546,483586,747623956,3,1.098612289,20.43241068
---
[2-2]上位７位：約全体(69銘柄)の１割？、R^2の１つ目の山。
2914,20081212,ＪＴ,311000,10000,3110000000,1,0,21.85788856
2503,20081212,キリン,1082,984508,1065237656,2,0.693147181,20.78646376
2502,20081212,アサヒ,1546,483586,747623956,3,1.098612289,20.43241068
2802,20081212,味の素,942,700033,659431086,4,1.386294361,20.30688803
2897,20081212,日清食品,3290,127464,419356560,5,1.609437912,19.85423209
2267,20081212,ヤクルト,1752,175910,308194320,6,1.791759469,19.54624105
2875,20081212,東洋水産,2645,110881,293280245,7,1.945910149,19.49663918
---
[2-3]上位１４位：とりあえず、７位の２倍ということで、、、：あまり意味はない。
2914,20081212,ＪＴ,311000,10000,3110000000,1,0,21.85788856
2503,20081212,キリン,1082,984508,1065237656,2,0.693147181,20.78646376
2502,20081212,アサヒ,1546,483586,747623956,3,1.098612289,20.43241068
2802,20081212,味の素,942,700033,659431086,4,1.386294361,20.30688803
2897,20081212,日清食品,3290,127464,419356560,5,1.609437912,19.85423209
2267,20081212,ヤクルト,1752,175910,308194320,6,1.791759469,19.54624105
2875,20081212,東洋水産,2645,110881,293280245,7,1.945910149,19.49663918
2212,20081212,山崎製パン,1284,220283,282843372,8,2.079441542,19.46040385
2282,20081212,日本ハム,1202,228445,274590890,9,2.197224577,19.43079288
2002,20081212,日清製粉,1078,251535,271154730,10,2.302585093,19.41820018
2579,20081212,北九州コカ,1942,111126,215806692,11,2.397895273,19.18989362
2801,20081212,キッコマン,978,210383,205754574,12,2.48490665,19.14219463
2501,20081212,サッポロ,507,393971,199743297,13,2.564949357,19.11254359
2809,20081212,キユーピー,1188,155465,184692420,14,2.63905733,19.0342024
---
[2-4]上位２８位：R^2の２つ目の山。R^2の最大値。
2914,20081212,ＪＴ,311000,10000,3110000000,1,0,21.85788856
2503,20081212,キリン,1082,984508,1065237656,2,0.693147181,20.78646376
2502,20081212,アサヒ,1546,483586,747623956,3,1.098612289,20.43241068
2802,20081212,味の素,942,700033,659431086,4,1.386294361,20.30688803
2897,20081212,日清食品,3290,127464,419356560,5,1.609437912,19.85423209
2267,20081212,ヤクルト,1752,175910,308194320,6,1.791759469,19.54624105
2875,20081212,東洋水産,2645,110881,293280245,7,1.945910149,19.49663918
2212,20081212,山崎製パン,1284,220283,282843372,8,2.079441542,19.46040385
2282,20081212,日本ハム,1202,228445,274590890,9,2.197224577,19.43079288
2002,20081212,日清製粉,1078,251535,271154730,10,2.302585093,19.41820018
2579,20081212,北九州コカ,1942,111126,215806692,11,2.397895273,19.18989362
2801,20081212,キッコマン,978,210383,205754574,12,2.48490665,19.14219463
2501,20081212,サッポロ,507,393971,199743297,13,2.564949357,19.11254359
2809,20081212,キユーピー,1188,155465,184692420,14,2.63905733,19.0342024
2810,20081212,ハウス食品,1586,110879,175854094,15,2.708050201,18.9851652
2202,20081212,明治製菓,408,385535,157298280,16,2.772588722,18.87365443
2261,20081212,明治乳業,466,329649,153616434,17,2.833213344,18.84996937
2811,20081212,カゴメ,1493,99617,148728181,18,2.890371758,18.81763091
2206,20081212,江崎グリコ,1010,144860,146308600,19,2.944438979,18.80122865
2871,20081212,ニチレイ,394,310851,122475294,20,2.995732274,18.62341989
2593,20081212,伊藤園,1260,91212,114927120,21,3.044522438,18.55980875
2607,20081212,不二製油,1242,87569,108760698,22,3.091042453,18.5046606
2531,20081212,宝ホールディ,488,217700,106237600,23,3.135494216,18.48118865
2262,20081212,雪印乳業,335,303802,101773670,24,3.17805383,18.43826198
2059,20081212,ユニチャーム,3250,29360,95420000,25,3.218875825,18.37379876
2264,20081212,森永乳業,345,253977,87622065,26,3.258096538,18.28854341
2602,20081212,日清オイリオ,471,173339,81642669,27,3.295836866,18.21786259
2001,20081212,日本製粉,461,174148,80282228,28,3.33220451,18.20105883
：売上高か？
：本に載っていない企業もあります、、、：？以下。
===
2531,20081212,宝ホールディ,488,217700,106237600,23,3.135494216,18.48118865
2059,20081212,ユニチャーム,3250,29360,95420000,25,3.218875825,18.37379876
===
~~~~~~~~~~~~~~~~~~~~~~~~~~~~(2008/12/19,15:10-)
[3]R^2最大値までの近似線を引いてみる。The approximation line to R^2 maximum value is drawn.
[3-1]全体を散布図でプロットする。
foods = read.csv("20081212-foods.txt");
plot(LN_jika ~ LN_order, data = foods)
---
[3-2]TOPから28位までの近似線をひく。
# TOPから28位までのデータセットを作成。
foods_LN_order_28 = foods$LN_order[1:28]
foods_LN_jika_28 = foods$LN_jika[1:28]
foods_28 = data.frame(LN_order=foods_LN_order_28, LN_jika=foods_LN_jika_28)
# データセットを単回帰分析。
# 線形近似を求める。
result = lm(LN_jika ~ LN_order, data = foods_28)
# 近似線をひく。
abline(result)
# 28位に垂直線をひく。
abline(v = log(28), col="red")
---
[3-3]3位、7位の近似線も重ねてみる。
foods_LN_order_07 = foods$LN_order[1:7]
foods_LN_jika_07 = foods$LN_jika[1:7]
foods_07 = data.frame(LN_order=foods_LN_order_07, LN_jika=foods_LN_jika_07)
result_07 = lm(LN_jika ~ LN_order, data = foods_07)
abline(result_07)
abline(v = log(7), col="red")
foods_LN_order_03 = foods$LN_order[1:3]
foods_LN_jika_03 = foods$LN_jika[1:3]
foods_03 = data.frame(LN_order=foods_LN_order_03, LN_jika=foods_LN_jika_03)
result_03 = lm(LN_jika ~ LN_order, data = foods_03)
abline(result_03)
abline(v = log(3), col="red")
~~~~~~~~~~~~~~~~~~~~~~~~~~~~(2008/12/19,15:10-)
[4]試行）２８位以降は、R^2の傾向はどうなっているか？
The trial) How does the tendency to R^2 become it since 28th place?
---
[4-1]２９位からラストまでで、R^2の状態を見る。
foods = read.csv("20081212-foods.txt");
kinji_r2 = c()
kinji_a = c()
kinji_b = c()
base_order = 29
###
search_order = base_order + 3
for (i in search_order:length(foods$LN_order)){
foods_LN_order_XX = foods$LN_order[base_order:i]
foods_LN_jika_XX = foods$LN_jika[base_order:i]
foods_XX = data.frame(LN_order=foods_LN_order_XX, LN_jika=foods_LN_jika_XX)
result = summary(result <- lm(LN_jika ~ LN_order, data = foods_XX))
kinji_r2[i] = result$r.squared
kinji_b[i] = result$coefficients[1]
kinji_a[i] = result$coefficients[2]
}
kinji_ALL = data.frame(KINJI_A=kinji_a, KINJI_B=kinji_b, KINJI_R2=kinji_r2)
#
plot(kinji_ALL$KINJI_R2)
---
> kinji_ALL
KINJI_A KINJI_B KINJI_R2
1 NA NA NA
2 NA NA NA
3 NA NA NA
4 NA NA NA
5 NA NA NA
6 NA NA NA
7 NA NA NA
8 NA NA NA
9 NA NA NA
10 NA NA NA
11 NA NA NA
12 NA NA NA
13 NA NA NA
14 NA NA NA
15 NA NA NA
16 NA NA NA
17 NA NA NA
18 NA NA NA
19 NA NA NA
20 NA NA NA
21 NA NA NA
22 NA NA NA
23 NA NA NA
24 NA NA NA
25 NA NA NA
26 NA NA NA
27 NA NA NA
28 NA NA NA
29 NA NA NA
30 NA NA NA
31 NA NA NA
32 -2.117902 25.12154 0.9469793
33 -1.903733 24.39334 0.9530634
34 -1.681958 23.63695 0.9415084
35 -1.747786 23.86214 0.9618844
36 -1.711026 23.73602 0.9714386
37 -1.820286 24.11197 0.9727569
38 -1.824618 24.12692 0.9795561
39 -1.763410 23.91513 0.9791543
40 -1.697247 23.68557 0.9766580
41 -1.638252 23.48034 0.9745409
42 -1.611233 23.38610 0.9770650
43 -1.575727 23.26195 0.9775093
44 -1.570542 23.24377 0.9807556
45 -1.545937 23.15731 0.9815584
46 -1.557878 23.19937 0.9839262
47 -1.592808 23.32271 0.9834323
48 -1.635288 23.47307 0.9814780
49 -1.679389 23.62952 0.9793602
50 -1.722936 23.78435 0.9775434
51 -1.746983 23.87004 0.9788431
52 -1.765034 23.93450 0.9803990
53 -1.794668 24.04056 0.9803327
54 -1.820187 24.13209 0.9807158
55 -1.895708 24.40352 0.9680491
56 -1.953377 24.61122 0.9634660
57 -2.015220 24.83441 0.9584019
58 -2.062529 25.00548 0.9574780
59 -2.104095 25.15609 0.9576190
60 -2.175617 25.41573 0.9503957
61 -2.229160 25.61049 0.9488734
62 -2.277106 25.78521 0.9485560
63 -2.315086 25.92388 0.9497910
64 -2.393985 26.21248 0.9413408
65 -2.489239 26.56153 0.9290079
66 -2.569910 26.85768 0.9234153
67 -2.661975 27.19626 0.9155266
68 -2.759161 27.55431 0.9075625
69 -2.873103 27.97481 0.8960784
>
---
[4-2]２９位からラストまでで、R^2の最大値を得る。
> max_order = which.max(kinji_ALL$KINJI_R2[32:69]) + 31
> max_r2 = kinji_ALL$KINJI_R2[max_order]
> max_order
[1] 46
> max_r2
[1] 0.9839262
>
---
[5]29位以降のデータに対して、46位の近似線を引く。
# 29位からラストまで、プロット。
foods = read.csv("20081212-foods.txt");
foods_LN_order_29toLast = foods$LN_order[29:length(foods$LN_order)]
foods_LN_jika_29toLast = foods$LN_jika[29:length(foods$LN_order)]
foods_29toLast = data.frame(LN_order=foods_LN_order_29toLast, LN_jika=foods_LN_jika_29toLast)
plot(LN_jika ~ LN_order, data = foods_29toLast)
# データセットを単回帰分析
result = lm(LN_jika ~ LN_order, data = foods_29toLast)
# 近似曲線をグラフに追加。
abline(result)
：この線は意味なし？？？
---
29位から46位の近似曲線では？
：17銘柄
start_order = 29
end_order = 46
foods_LN_order_start2end = foods$LN_order[start_order:end_order]
foods_LN_jika_start2end = foods$LN_jika[start_order:end_order]
foods_start2end = data.frame(LN_order=foods_LN_order_start2end, LN_jika=foods_LN_jika_start2end)
result = lm(LN_jika ~ LN_order, data = foods_start2end)
abline(result)
abline(v = log(end_order), col="red")
~~~~~~~~~~~~~~~~~~~~~~~~~~~~(2008/12/19,15:10-)
[5]出来高と順位の関係をチェックする。
The relation between the trading volume and the order is checked.
：出来高は何位までの集中しているか？
On how much order has the trading volume concentrated?
---
2008/12/12の出来高データ(dekidaka)は以下。：東証一部の食品。
===
code,name,owarine,stock_num,jika,order,dekidaka
2914,ＪＴ,311000,10000,3110000000,1,40.079
2503,キリン,1082,984508,1065237656,2,7134
2502,アサヒ,1546,483586,747623956,3,4101.6
2802,味の素,942,700033,659431086,4,7837
2897,日清食品,3290,127464,419356560,5,366.9
2267,ヤクルト,1752,175910,308194320,6,724.4
2875,東洋水産,2645,110881,293280245,7,1120
2212,山崎製パン,1284,220283,282843372,8,781
2282,日本ハム,1202,228445,274590890,9,3780
2002,日清製粉,1078,251535,271154730,10,3312.5
2579,北九州コカ,1942,111126,215806692,11,647.8
2801,キッコマン,978,210383,205754574,12,3356
2501,サッポロ,507,393971,199743297,13,3154
2809,キユーピー,1188,155465,184692420,14,585.9
2810,ハウス食品,1586,110879,175854094,15,390.4
2202,明治製菓,408,385535,157298280,16,3738
2261,明治乳業,466,329649,153616434,17,3183
2811,カゴメ,1493,99617,148728181,18,267.4
2206,江崎グリコ,1010,144860,146308600,19,202
2871,ニチレイ,394,310851,122475294,20,3574
2593,伊藤園,1260,91212,114927120,21,363.7
2607,不二製油,1242,87569,108760698,22,343.5
2531,宝ホールディ,488,217700,106237600,23,3097
2262,雪印乳業,335,303802,101773670,24,2405.5
2059,ユニチャーム,3250,29360,95420000,25,70.3
2264,森永乳業,345,253977,87622065,26,865
2602,日清オイリオ,471,173339,81642669,27,2389
2001,日本製粉,461,174148,80282228,28,690
2284,伊藤ハム,315,210483,66302145,29,863
2580,コカコーラセ,654000,90,58860000,30,0.142
2613,Ｊ－オイルミ,335,167542,56126570,31,1042
2004,昭和産業,296,180650,53472400,32,257
2201,森永製菓,189,270949,51209361,33,535
2109,三井製糖,354,141667,50150118,34,346
2572,三国コカ,853,53556,45683268,35,103.3
2815,アリアケ,1359,32809,44587431,36,102.8
2207,名糖産業,1884,21265,40063260,37,13.2
2590,ダイドードリ,2375,16569,39351375,38,49.3
2908,フジッコ,1120,34992,39191040,39,100
2281,プリマハム,171,224393,38371203,40,2761
2108,日本甜菜糖,244,153256,37394464,41,445
2288,丸大食品,268,132528,35517504,42,362
2594,キーコーヒ,1554,22464,34909056,43,30.8
2899,永谷園,857,38277,32803389,44,29
2290,米久,1134,28810,32670540,45,85
2204,中村屋,503,59762,30060286,46,107
2540,養命酒製造,840,33000,27720000,47,27
2918,わらべや日洋,1561,16626,25953186,48,51.4
2536,メルシャン,183,133689,24465087,49,220
2211,不二家,119,194377,23130863,50,259
2292,ＳＦＯＯＤＳ,711,32268,22942548,51,19.5
2051,日本農産工,172,129310,22241320,52,143
2578,四国コカ,860,23908,20560880,53,25.8
2009,鳥越製粉,760,26036,19787360,54,38.4
2053,中部飼料,599,26536,15895064,55,65
2910,ロックフィ,1166,13394,15617404,56,23.9
2533,オエノンホー,218,65586,14297748,57,630
2812,焼津水産化学,1000,14056,14056000,58,19.6
2003,日東製粉,287,46924,13467188,59,24
2922,なとり,709,15532,11012188,60,21.4
2217,モロゾフ,300,36692,11007600,61,59
2052,協同飼料,101,103996,10503596,62,177
4404,ミヨシ油脂,125,82455,10306875,63,158
2597,ユニカフェ,1118,6869,7679542,64,10.5
2286,林兼産業,71,89100,6326100,65,316
2056,日配合飼料,86,71877,6181422,66,472
2107,東洋精糖,95,54560,5183200,67,417
2215,第一パン,93,48048,4468464,68,59
2599,ジャパンフー,691,5100,3524100,69,8.5
===
出来高全体：68996.521 単位：千株数
---
３位 ：１６％ =(11275.679/68996.521)*100=16.34238776
７位 ：３０％ =(21323.979/68996.521)*100=30.90587567
２８位 ：８４％ =(58519.979/68996.521)*100=84.81584021
４６位 ：９５％ =(65751.521/68996.521)*100=95.29686432
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[6]一時的なまとめ
トップからの順位でR^2の最大値は、今回のデータで２８位。
トップから２８位までの出来高が８４％を占めることから、単純なR^2の最大値で近似しても、
マーケットの動きをカバーできるのではないか、、、
２９位から４６位までの銘柄は、１位から２８位の動きに引きずられるようなものか？？？
---
当面のターゲットは、トップからのR^2で最大値までの範囲をべき乗傾向の範囲とする。
：これは、Excelでの手作業と結果的には同じになった。
~~~
番外）
筆記試験の準備のため。exit。遅い？
：(３日前)本屋の採用試験コーナーに行く。驚く！SPI...WEB：採用試験もインターネットの時代か、、、
~~~
end
0 件のコメント:
コメントを投稿