Another important recommendation from our developers is to contact Microsoft directly. In other words, once any customer manages to reproduce the issue while collecting the requested log they should raise a ticket with Microsoft. Without analysis from MS we won’t be able to do anything on our side, so involving MS is crucial at this point. To speed things up, the customer should refer to advisory case # 117062115926216.
2019年6月26日水曜日
Windows profile bug
Another important recommendation from our developers is to contact Microsoft directly. In other words, once any customer manages to reproduce the issue while collecting the requested log they should raise a ticket with Microsoft. Without analysis from MS we won’t be able to do anything on our side, so involving MS is crucial at this point. To speed things up, the customer should refer to advisory case # 117062115926216.
2019年6月19日水曜日
EPS 2019JUN19
> eps_year_xts["2019::"]
[,1]
2019-01-01 134.48
2019-04-01 137.14
2019-07-01 139.58
2019-10-01 150.77
2020-01-01 153.98
2020-04-01 158.28
2020-07-01 162.93
2020-10-01 167.58
> eps_year_xts["2019::"][1]
[,1]
2019-01-01 134.48
> eps_year_xts["2019::"][1] <- 134.32
> eps_year_xts["2019::"][2] <- 137.01
> eps_year_xts["2019::"][3] <- 139.53
> eps_year_xts["2019::"][4] <- 150.72
> eps_year_xts["2019::"][5] <- 154.31
> eps_year_xts["2019::"][6] <- 158.70
> eps_year_xts["2019::"][7] <- 163.52
> eps_year_xts["2019::"][8] <- 168.46
> eps_year_xts["2019::"]
[,1]
2019-01-01 134.32
2019-04-01 137.01
2019-07-01 139.53
2019-10-01 150.72
2020-01-01 154.31
2020-04-01 158.70
2020-07-01 163.52
2020-10-01 168.46
eps_year_xts["2019::"][1] <- 134.32
eps_year_xts["2019::"][2] <- 137.01
eps_year_xts["2019::"][3] <- 139.53
eps_year_xts["2019::"][4] <- 150.72
eps_year_xts["2019::"][5] <- 154.31
eps_year_xts["2019::"][6] <- 158.70
eps_year_xts["2019::"][7] <- 163.52
eps_year_xts["2019::"][8] <- 168.46
$cat eps.txt
12/31/2020 $49.27 $45.10 15.65 17.17 $184.82 $168.46
9/30/2020 $47.66 $43.70 16.09 17.68 $179.67 $163.52
6/30/2020 $45.31 $41.13 16.55 18.22 $174.77 $158.70
3/31/2020 $42.58 $38.53 17.06 18.74 $169.53 $154.31
12/31/2019 $44.12 $40.16 17.53 19.19 $164.93 $150.72
9/30/2019 $42.76 $38.88 18.56 20.72 $155.84 $139.53
6/30/2019 $40.07 $36.74 18.72 21.11 $154.46 $137.01
3/31/2019 (98.2%) 2834.40 $37.98 $34.95 18.89 21.53 $153.04 $134.32
のとき
$tac eps.txt | awk '{gsub("\\$","",$NF);print "eps_year_xts[\"2019::\"]["NR"] <- "$NF}'
eps_year_xts["2019::"][1] <- 134.32
eps_year_xts["2019::"][2] <- 137.01
eps_year_xts["2019::"][3] <- 139.53
eps_year_xts["2019::"][4] <- 150.72
eps_year_xts["2019::"][5] <- 154.31
eps_year_xts["2019::"][6] <- 158.70
eps_year_xts["2019::"][7] <- 163.52
eps_year_xts["2019::"][8] <- 168.46
2019年6月18日火曜日
サトウキビ(砂糖キビ) -- sugarcane.
from 2019/3/26 conversation.
Cost for 3 acre plowing and planting.
- sugarcane varieties
- worker wage,plow and planning
- Fertilizer packages price = 40,000 baht for 1 acre.. ????
- Food pice 10,000 baht...
Harvest
- 3 acre = 90 ton
- selling price is 1 ton =900 baht.
- 90 ton * 900 bhat = 81,000 baht
- if add 2 more acre, 2 acre = 60 ton = 54,000 THB
- total 5 acre = 150 ton = 135,000 THB
- 5 acre = 2.02343 ha.
Cost 2 acre to plow and plant
- 30,000 baht..
買取価格
上記「Harvest」と値が違うので注意。
- 自分で収穫、工場に納める場合 700 thb/t
- 収穫、輸送を依頼する場合 500 thb/t
2019年6月17日月曜日
CLI - composite leading indicator - USA, China and Europe Area
> summary(lm(cli_xts$oecd ~ cli_xts$usa + cli_xts$ea19 + cli_xts$china))
Call:
lm(formula = cli_xts$oecd ~ cli_xts$usa + cli_xts$ea19 + cli_xts$china)
Residuals:
Min 1Q Median 3Q Max
-0.68475 -0.09634 0.03744 0.11972 0.41636
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 9.265072 1.036438 8.939 <2e-16 ***
cli_xts$usa 0.402771 0.013317 30.246 <2e-16 ***
cli_xts$ea19 0.438171 0.012302 35.618 <2e-16 ***
cli_xts$china 0.065993 0.006695 9.857 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 0.1958 on 348 degrees of freedom
(420 observations deleted due to missingness)
Multiple R-squared: 0.9577, Adjusted R-squared: 0.9573
F-statistic: 2624 on 3 and 348 DF, p-value: < 2.2e-16
> last(cli_xts,n=12)
oecd usa china ea19
2018-05-01 100.33250 100.47600 99.29018 100.53100
2018-06-01 100.22810 100.42840 99.13142 100.40680
2018-07-01 100.11390 100.36340 98.96715 100.28170
2018-08-01 99.98618 100.27230 98.80726 100.14970
2018-09-01 99.84174 100.13410 98.67062 100.01150
2018-10-01 99.68746 99.94146 98.57309 99.87226
2018-11-01 99.53398 99.71795 98.51999 99.73539
2018-12-01 99.38741 99.48866 98.50362 99.59542
2019-01-01 99.25877 99.28568 98.52641 99.45764
2019-02-01 99.15687 99.12820 98.58793 99.32590
2019-03-01 99.08728 99.01845 98.69128 99.20160
2019-04-01 99.03429 98.93993 98.80725 99.08523
for future reference.
> last(tmp.predict,n=6)
SP5.Open SP5.High SP5.Low SP5.Close SP5.Volume spline eps
1 2019 2476.96 2708.95 2443.96 2704.10 80391630000 2906.849 2760.208
2 2019 2702.32 2813.49 2681.83 2784.49 70183430000 2929.656 2779.348
3 2019 2798.22 2860.31 2722.27 2834.40 78596280000 2956.669 2804.840
4 2019 2848.63 2949.52 2848.63 2945.83 69604840000 2983.000 2829.000
5 2019 2952.33 2954.13 2750.52 2752.06 76860120000 3010.000 2854.000
6 2019 2751.53 2910.61 2728.81 2886.98 33703630000 3037.000 2879.000
2019年6月3日月曜日
EPS 2019JUN03
Download XLS from here.
> eps_year_xts["2019::"]
[,1]
2019-01-01 134.74
2019-04-01 137.38
2019-07-01 140.00
2019-10-01 151.44
2020-01-01 154.73
2020-04-01 159.04
2020-07-01 163.53
2020-10-01 167.98
> eps_year_xts["2019-01-01"] <- 134.48
> eps_year_xts["2019-04-01"] <- 137.14
> eps_year_xts["2019-07-01"] <- 139.58
> eps_year_xts["2019-10-01"] <- 150.77
> eps_year_xts["2020-01-01"] <- 153.98
> eps_year_xts["2020-04-01"] <- 158.28
> eps_year_xts["2020-07-01"] <- 162.93
> eps_year_xts["2020-10-01"] <- 167.58
> eps_year_xts["2019::"]
[,1]
2019-01-01 134.48
2019-04-01 137.14
2019-07-01 139.58
2019-10-01 150.77
2020-01-01 153.98
2020-04-01 158.28
2020-07-01 162.93
2020-10-01 167.58
eps_year_xts["2019-01-01"] <- 134.48
eps_year_xts["2019-04-01"] <- 137.14
eps_year_xts["2019-07-01"] <- 139.58
eps_year_xts["2019-10-01"] <- 150.77
eps_year_xts["2020-01-01"] <- 153.98
eps_year_xts["2020-04-01"] <- 158.28
eps_year_xts["2020-07-01"] <- 162.93
eps_year_xts["2020-10-01"] <- 167.58
histogram performance comparison between cli 1 month delta positive and negative.
when "func()" is "cli_delta_vs_period_return_rate.r"
> hist(as.vector(func("minus","1970-01-01")[,1])-1,col=rgb(0.5,1,0),breaks=20,xlim=c(-0.6,0.5),ylim=c(0,7))
0111 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
[1] 11.34615
[1] 0.9858096
> par(new=T)
> hist(as.vector(func("plus","1970-01-01")[,1])-1,col=rgb(0.5,0,1,alpha=0.4),breaks=10,xlim=c(-0.6,0.5),ylim=c(0,7))
source code is as below.
#
# 1)pick up months whose clie_xts$oecd is up from the previous
# 2)create the stream of flags, in which up is 1 and down is 0
# 3)compare SPX close price between the end month of cli_xts delta is plus(or minus) and its start.
# 4)return xts objects which contains. start month of period, updown ration, length of months and monthly average return during period.
func <- function(pm="plus",s="1970-01-01",l=1){
w <- c()
cat("0")
cat(length(s))
if(nchar(s) == 10){ # use nchar() to measure strings length, not length()
cat("1")
last_date <- last(index(cli_xts$oecd))
start_date <- s
period <- paste(start_date,last_date,sep='::')
}else{
# last_date <- l
period <- s
}
# last_date <- last(index(cli_xts$oecd))
# start_date <- s
# period <- paste(start_date,last_date,sep='::')
start_index <- 1
iteration <- 0
performance_val <- c()
period_length <- c()
lag_month <- l
result <- c()
rate <- c()
plus_or_minus <- pm
open_p <- c()
close_p <- c()
# put flag on the months accoring to the parameter. for "minus" cli delta is less than ZERO, for plus the opposite.
for(i in seq(1,length(diff(cli_xts$oecd,lag=lag_month)[period]),1,)){
if(plus_or_minus == "minus"){
if(as.vector(diff(cli_xts$oecd,lag=lag_month)[period])[i] < 0){ # up is "> 0"
w <- append(w,1)
}else{
w <- append(w,0)
}
}else if(plus_or_minus == "plus"){
if(as.vector(diff(cli_xts$oecd,lag=lag_month)[period])[i] > 0){ # up is "> 0"
w <- append(w,1)
}else{
w <- append(w,0)
}
}else{
stop("please use plus or minus as 1st parameter")
}
}
month_flag <- 0 # status flag
# check stream and when flag is changes 0 to 1. it is the start of period.
for(i in seq(1,length(diff(cli_xts$oecd,lag=lag_month)[period]),1,)){
if(w[i] == 1){
if(month_flag == 0){ # when w is 1 and month_flag is 0, the period starts
month_flag <- 1
start_price <- as.vector(to.monthly(SP5[period])[,1][i]) #dc 0531
# print(index(to.monthly(SP5[period])[,4][i]))
# cat("from ")
# cat(as.character(as.Date(index(to.monthly(SP5[period])[,4][i]))))
start_index <- i
}
# dc 0602 add output at the end of loop
if(i == length(diff(cli_xts$oecd,lag=lag_month)[period])){
# print("end of the loop")
result <- append(result,as.xts(as.vector(to.monthly(SP5[period])[,4][i]) / start_price,index(to.monthly(SP5[period])[,4][i])))
period_length <- append(period_length,i-start_index)
performance_val <- append(performance_val,as.vector(to.monthly(SP5[period])[,4][i-1]) / start_price)
open_p <- append(open_p,start_price)
close_p <- append(close_p,as.vector(to.monthly(SP5[period])[,4][i-1]))
rate <- append(rate,last(performance_val)**(1/last(period_length))-1)
}
# check stream and when flag is changes 1 to 0. it is the start of period.
}else if(w[i] ==0){
if(month_flag == 1){ # when w is 0 and month_flag is 1, the period ends
# cat(" ")
# cat(i - start_index)
# cat(" month(s)")
# cat("\n")
iteration <- iteration +1
cat(iteration)
cat(" ")
result <- append(result,as.xts(as.vector(to.monthly(SP5[period])[,4][i-1]) / start_price,index(to.monthly(SP5[period])[,4][i-1])))
# print(as.xts(as.vector(to.monthly(SP5[period])[,4][i]) / start_price,index(to.monthly(SP5[period])[,4][i])))
# print(i - start_index)
month_flag <- 0 # when period ends, intialize the flag.
period_length <- append(period_length,i-start_index)
performance_val <- append(performance_val,as.vector(to.monthly(SP5[period])[,4][i-1]) / start_price)
open_p <- append(open_p,start_price)
close_p <- append(close_p,as.vector(to.monthly(SP5[period])[,4][i-1]))
rate <- append(rate,last(performance_val)**(1/last(period_length))-1)
}
}
}
cat("\n")
print(mean(period_length))
print(mean(performance_val))
return(merge(result,period_length,rate,open_p,close_p))
}
# t_minus <- performance_val
# t_plus <- performance_val
func("minus","1970-01-01")
2019年6月2日日曜日
S&P 500 performance comparison between CLI 1 month delta is positive and negative.
The period when OECD Composite Leading Inditacor 1 month delta is positive and negative come one after another. Here to calculate S&P 500 return during those each period.
The parameter "minus" indicates this is period when CLI delta is negative. The result is they came 25 times since 1970/1/1. The average length is approx. 11.35 month and the average return is -1.41%.
On the other hand, positive have come 25 times. The average length is 11.8 month and the average return is 19.85%.
> func("minus","1970-01-01")
0111 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
[1] 11.34615
[1] 0.9858096
result period_length rate open_p close_p
Sep 1970 0.9157072 9 -0.0097365802 92.06 84.30
Jan 1975 0.6634491 24 -0.0169506551 116.03 76.98
Jul 1977 0.9867239 14 -0.0009541901 100.18 98.85
Jun 1980 1.2136407 20 0.0097282501 94.13 114.24
Aug 1982 0.8803035 20 -0.0063541523 135.76 119.51
Nov 1984 1.0010403 10 0.0001039829 163.41 163.58
Jul 1986 1.3032344 17 0.0157013530 181.18 236.12
Mar 1988 0.7849672 7 -0.0339963144 329.81 258.89
Oct 1989 1.2255509 10 0.0205472899 277.72 340.36
Jan 1991 1.0362770 11 0.0032447523 331.89 343.93
Dec 1991 1.0547758 4 0.0134213316 395.43 417.09
Nov 1992 1.0395228 7 0.0055527568 414.95 431.35
Jun 1995 1.1773542 9 0.0183066251 462.69 544.75
Jan 1996 1.0937576 3 0.0303236982 581.50 636.02
Oct 1998 1.1598155 13 0.0114699680 947.28 1098.67
Sep 2001 0.6946176 18 -0.0200405678 1498.58 1040.94
Mar 2003 0.7875979 11 -0.0214722593 1076.92 848.18
May 2005 1.0579732 14 0.0040334726 1126.21 1191.50
Jun 2006 1.0001180 1 0.0001180284 1270.05 1270.20
Feb 2009 0.4885423 20 -0.0351826442 1504.66 735.09
Jun 2010 0.9479536 1 -0.0520464319 1087.30 1030.71
Nov 2011 0.9385236 9 -0.0070249109 1328.64 1246.96
Sep 2012 1.0547405 7 0.0076425915 1365.90 1440.67
Aug 2014 1.0690570 5 0.0134449700 1873.96 2003.37
Apr 2016 1.0031085 16 0.0001940004 2058.90 2065.30
Mar 2019 1.0599561 15 0.0034295890 2645.10 2784.49
> func("plus","1970-01-01")
0111 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
[1] 11.8
[1] 1.198459
result period_length rate open_p close_p
Jan 1973 1.3763938 28 0.011474868 84.30 116.03
May 1976 1.3013769 16 0.016600207 76.98 100.18
Oct 1978 0.9423369 15 -0.003951666 98.85 93.15
Dec 1980 1.1883753 6 0.029182211 114.24 135.76
Jan 1984 1.3672189 17 0.018569047 119.52 163.41
Feb 1985 1.1075926 3 0.034649717 163.58 181.18
Aug 1987 1.3967474 13 0.026036742 236.12 329.80
Dec 1988 1.0727335 9 0.007831630 258.89 277.72
Feb 1990 0.9751147 4 -0.006280247 340.36 331.89
Aug 1991 1.1498066 7 0.020142134 343.91 395.43
Apr 1992 0.9950124 4 -0.001249244 417.03 414.95
Sep 1994 1.0727019 22 0.003195123 431.35 462.71
Oct 1995 1.0674621 4 0.016454915 544.75 581.50
Sep 1997 1.4893871 20 0.020117927 636.02 947.28
Mar 2000 1.3639946 17 0.018427587 1098.67 1498.58
Apr 2002 1.0345650 7 0.004866239 1040.94 1076.92
Mar 2004 1.3277960 12 0.023908021 848.18 1126.21
May 2006 1.0659588 12 0.005337085 1191.50 1270.09
Jun 2007 1.1836842 12 0.014151848 1270.06 1503.35
May 2010 1.4932221 15 0.027089509 729.57 1089.41
Feb 2011 1.2871884 8 0.032060761 1031.10 1327.22
Feb 2012 1.0952515 3 0.030792577 1246.91 1365.68
Mar 2014 1.2994239 18 0.014657552 1440.90 1872.34
Dec 2014 1.0273593 4 0.006770750 2004.07 2058.90
Nov 2017 1.2807753 19 0.013109691 2067.17 2647.58
their results are tested by "t.test()". p-value is 7.391e-05. Therefore they should be distiguised each other with the probability of 99.99261%
> t.test(func("plus","1970-01-01")[,1],func("minus","1970-01-01")[,1])
<skip>
Welch Two Sample t-test
data: func("plus", "1970-01-01")[, 1] and func("minus", "1970-01-01")[, 1]
t = 4.3306, df = 48.714, p-value = 7.391e-05
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
0.1138069 0.3109339
sample estimates:
mean of x mean of y
1.1984592 0.9860888
There is a histogram in the next entry for the better comparison.
see source code github. commit id is d20207a
登録:
投稿 (Atom)