Presentations Discussion
JestonBlu opened this issue · comments
Anything specific to the in class presentations can go here
For the first presentation, I am going to start organizing here: https://www.overleaf.com/5451263qytmjt.
Anyone can edit, its on overleaf which is like github for LaTex.
Nice job. We could also put a link some where to our github repository to show how we are managing the project.
Good idea.
On Sun, Jun 12, 2016 at 11:38 AM, Joseph Blubaugh notifications@github.com
wrote:
Nice job. We could also put a link some where to our github repository to
show how we are managing the project.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D225452991&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=5wtEHP-YHg_zDy_PI4E-m5SiRBBmQkHpHFepeWsrM6o&s=c7MT6zTbqR9KbBLl7nrLd5EsXld5xQ9T3SKzDky5YDc&e=,
or mute the thread
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK01LqhFytxXdXDFz2wPe9CCHycmPiks5qLFIcgaJpZM4Izj99&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=5wtEHP-YHg_zDy_PI4E-m5SiRBBmQkHpHFepeWsrM6o&s=ZyjDD7_qLmV1FxrIWTmu35X2wxMLmcPG6ez6UG0At1g&e=
.
Hey guys,
I will start with the analysis and plotting today at my end. And also, I
can write the final project report.
Regards,
Akarshan
On Jun 12, 2016 2:20 PM, "Alison" notifications@github.com wrote:
Good idea.
On Sun, Jun 12, 2016 at 11:38 AM, Joseph Blubaugh <
notifications@github.com>
wrote:Nice job. We could also put a link some where to our github repository to
show how we are managing the project.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D225452991&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=5wtEHP-YHg_zDy_PI4E-m5SiRBBmQkHpHFepeWsrM6o&s=c7MT6zTbqR9KbBLl7nrLd5EsXld5xQ9T3SKzDky5YDc&e=
,
or mute the thread
<
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK01LqhFytxXdXDFz2wPe9CCHycmPiks5qLFIcgaJpZM4Izj99&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=5wtEHP-YHg_zDy_PI4E-m5SiRBBmQkHpHFepeWsrM6o&s=ZyjDD7_qLmV1FxrIWTmu35X2wxMLmcPG6ez6UG0At1g&e=.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AS9ZZ0yvJBzSFnoTbVD17ZPsdU7h6-Upks5qLFwPgaJpZM4Izj99
.
Akarshan, make sure you look at the github project main page to see who is
working on what, i dont think we want to replicate each others work so make
sure someone else isn't already doing what you want to do.. for example I
have already posted some plots and data prep, and sean posted acf/pacf
plots as well... Alison is already working on the presentation and signed
up for the write up early so make sure you coordinate with her if you want
to be involved in that part. Travis has posted that he is interested in
diagnostics so if you want to work on the modeling aspect coordinate with
him on that.
Also please fill in next to your name on the readme document, what you are
working on so everyone knows. Same for you Bo, please fill in what you are
working on
On Sun, Jun 12, 2016 at 2:23 PM, pakarshan notifications@github.com wrote:
Hey guys,
I will start with the analysis and plotting today at my end. And also, I
can write the final project report.Regards,
Akarshan
On Jun 12, 2016 2:20 PM, "Alison" notifications@github.com wrote:Good idea.
On Sun, Jun 12, 2016 at 11:38 AM, Joseph Blubaugh <
notifications@github.com>
wrote:Nice job. We could also put a link some where to our github repository
to
show how we are managing the project.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<,
or mute the thread
<.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<
#5 (comment)
,
or mute the thread
<
https://github.com/notifications/unsubscribe/AS9ZZ0yvJBzSFnoTbVD17ZPsdU7h6-Upks5qLFwPgaJpZM4Izj99.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/ADL2haJ8NDHkEUOt_zL095yr99C5zpcCks5qLFyUgaJpZM4Izj99
.
Sure .. sounds good.
Ill coordinate with alison and travis.
On Jun 12, 2016 2:31 PM, "Joseph Blubaugh" notifications@github.com wrote:
Akarshan, make sure you look at the github project main page to see who is
working on what, i dont think we want to replicate each others work so make
sure someone else isn't already doing what you want to do.. for example I
have already posted some plots and data prep, and sean posted acf/pacf
plots as well... Alison is already working on the presentation and signed
up for the write up early so make sure you coordinate with her if you want
to be involved in that part. Travis has posted that he is interested in
diagnostics so if you want to work on the modeling aspect coordinate with
him on that.Also please fill in next to your name on the readme document, what you are
working on so everyone knows. Same for you Bo, please fill in what you are
working onTha
On Sun, Jun 12, 2016 at 2:23 PM, pakarshan notifications@github.com
wrote:Hey guys,
I will start with the analysis and plotting today at my end. And also, I
can write the final project report.Regards,
Akarshan
On Jun 12, 2016 2:20 PM, "Alison" notifications@github.com wrote:Good idea.
On Sun, Jun 12, 2016 at 11:38 AM, Joseph Blubaugh <
notifications@github.com>
wrote:Nice job. We could also put a link some where to our github
repository
to
show how we are managing the project.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<,
or mute the thread
<.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<,
or mute the thread
<https://github.com/notifications/unsubscribe/AS9ZZ0yvJBzSFnoTbVD17ZPsdU7h6-Upks5qLFwPgaJpZM4Izj99
.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<
#5 (comment)
,
or mute the thread
<
https://github.com/notifications/unsubscribe/ADL2haJ8NDHkEUOt_zL095yr99C5zpcCks5qLFyUgaJpZM4Izj99.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AS9ZZ7MhCctwJubc4-HXOXGQuq4vgWjZks5qLF51gaJpZM4Izj99
.
Alison, do you know what dimensions will look best on beemer? I can resize the plots Im generating, just let me know.
I signed up some responsibilities although they are flexible.
On Sun, Jun 12, 2016 at 2:31 PM, Joseph Blubaugh notifications@github.com
wrote:
Akarshan, make sure you look at the github project main page to see who is
working on what, i dont think we want to replicate each others work so make
sure someone else isn't already doing what you want to do.. for example I
have already posted some plots and data prep, and sean posted acf/pacf
plots as well... Alison is already working on the presentation and signed
up for the write up early so make sure you coordinate with her if you want
to be involved in that part. Travis has posted that he is interested in
diagnostics so if you want to work on the modeling aspect coordinate with
him on that.Also please fill in next to your name on the readme document, what you are
working on so everyone knows. Same for you Bo, please fill in what you are
working onTha
On Sun, Jun 12, 2016 at 2:23 PM, pakarshan notifications@github.com
wrote:Hey guys,
I will start with the analysis and plotting today at my end. And also, I
can write the final project report.Regards,
Akarshan
On Jun 12, 2016 2:20 PM, "Alison" notifications@github.com wrote:Good idea.
On Sun, Jun 12, 2016 at 11:38 AM, Joseph Blubaugh <
notifications@github.com>
wrote:Nice job. We could also put a link some where to our github
repository
to
show how we are managing the project.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<,
or mute the thread
<.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<,
or mute the thread
<https://github.com/notifications/unsubscribe/AS9ZZ0yvJBzSFnoTbVD17ZPsdU7h6-Upks5qLFwPgaJpZM4Izj99
.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<
#5 (comment)
,
or mute the thread
<
https://github.com/notifications/unsubscribe/ADL2haJ8NDHkEUOt_zL095yr99C5zpcCks5qLFyUgaJpZM4Izj99.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AKL-etucBCLGGBRMuGTQYpRWNfY0znR-ks5qLF51gaJpZM4Izj99
.
Beemer is pdf presentation, so each slide is letter sized.
As long as you post your code as well, it is easy to adjust dimensions if they are not working well.
I am not sure how to make this on overleaf, but I think there should be a second slide "Data" that contains definitions of our response and predictors:
For instance:
Unemployment: percentage of the labor force considered unemployed; Labor force includes
people at least 16 years of age who reside in the US but not institutions, and who are not in the Armed Forces (Federal Reserve Bank of St. Louis)
I will continue to look for definitions for the other predictors.
I'm on it. From what I saw, the frame environment does the job.
\begin{frame}
...foo...
\end{frame}
On Jun 12, 2016 4:22 PM, "trlilley12" notifications@github.com wrote:
I am not sure how to make this on overleaf, but I think there should be a
second slide "Data" that contains definitions of our response and
predictors:For instance:
Unemployment: percentage of the labor force considered unemployed; Labor
force includes
people at least 16 years of age who reside in the US but not institutions,
and who are not in the Armed Forces (Federal Reserve Bank of St. Louis)I will continue to look for definitions for the other predictors.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AS8he_c3Gj_Ew-fkwQRUn8BvqgK1xj6dks5qLHixgaJpZM4Izj99
.
Yep, all of the plot code is stored in RScripts/Exploratory_Plots.r
Industrial Production Index: output of manufacturing, mining, and electric and gas utilities industries, expressed as an index relative to an arbitrary base year of 2012; for example, an index of 104 means 4% greater output than in 2012 (Board of Governors of Federal Reserve System)
Purchase Only House Price Index: measure of single family house prices as an index relative to an arbitrary month of Jan 1991; meaning of index same as that for IPI (Federal Housing Finance Agency)
I think the other predictors are mostly self-explanatory and are measured in millions of dollars (Federal Reserve Bank of St. Louis)
In the Data/ folder there is an excel workbook that has the original data I compiled... the first several rows have the official titles of the data and their definitions from the source... use that rather than searching the web for the definitions
I looked at the ones from the Excel sheet, but they were very long and not exactly straightforward. I think we are supposed to assume our audience has just a basic background in econ.
Yeah, i would paraphrase what is on the excel sheet, there are 2 indicators that are indexes, the other 3 are in dollars... I created a data prep script and ended up scaling the dollar variables, but im not showing that on any of the plots so far
I've got only two of the variables explained so far; I can't find a short descriptor of the New Manufacturing Orders and Purchase Only House Price Index variables.
If you look above, I have the definition of HPI. New Manufacturing Orders is the amount (in millions) spend on manufacturing orders. I don't think it needs to be more complicated than that.
I think we should include only (1) the scatter plot of the predictors and (2) the time series plots for all variables in our 5 min presentation. I would add them myself, but I think I am causing problems whenever I try to edit overleaf.
The plot of unemployment overlaying the different presidents is very interesting, but I don't think it's relevant to our project. If we want to assess presidential impact on unemployment, we would need to introduce president as a categorical predictor, which I think is too complicated. But if everyone else wants to include the president, the dominant party in Congress, and an indicator for recession, we can do that.
I also think we should expand a little bit on what exactly unemployment is, since it is our main concern.
The Bureau of Labor Statistics defines someone in the labor force as unemployed if they are actively seeking a job, jobless, or available to take a job. By contrast, anyone with even part-time employment is considered employed.
The excel file has the official description and the source of the data so
maybe you could just paraphrase that if you dont want to use the official
title. All of the data came from FRED and the data set ID is on the second
row so you should be able to use that to find the webpage it came from.
New Manufacturing Orders is just the dollar value in millions of all new
orders placed in the current month. House Price Index is that change in
Single Family Housing Prices with 1991 being the base year and equal to
100.
On Sun, Jun 12, 2016 at 5:01 PM, SZRoberson notifications@github.com
wrote:
I've got only two of the variables explained so far; I can't find a short
descriptor of the New Manufacturing Orders and Purchase Only House Price
Index variables.—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/ADL2hQztj95GhQQAPxoHtBpw-oYa6AAiks5qLIGrgaJpZM4Izj99
.
Here is where the presentation is so far. Does anyone have any suggested changes?
Sean, do you want words with the plots or to explain them as is. Also, when I use beemer I usually open it in a pdf reader and click on view full screen. (That way I can arrow forward through the presentation and the text changes look like animations.)
Anything specific we want people to notice about the graphs? Is the resolution OK with everyone? I was going to redo them but then it looked fine to me on pdf full screen.
stat626projectInitialPres.pdf
I think its looking good.. My original thought of collecting the political data was to bring some non economic context to some of the graphs, maybe give an extra talking point or two since we one of the grading aspects of this project is how interesting the material is... i hadn't really thought about using the political data as prediction variables.. what does everyone else think? If we have enough people dedicated to modeling I think it would be interesting to explore.
It looks good, just made a minor fix.
I'll probably have index cards with the meaning of the plots written, since I think having words on the slides clutters it a little.
Looks good. Thanks for the nice job.
I just added model fitting to responsibility.
One minor question is whether we should present a few simple models for the first presentation. Just not sure if 5 mins would be enough.
I think we will learn more in todays lecture, but I dont think model fitting is something that will be discussed in the first presentations... i think its basically an overview of the data and some plots and thats about it. Thats about all you can do in 5 minutes anway.
I agree. Then everything for the first presentation looks good now.
Awesome. Please let me know if you need anything. I am in the middle of class right now but i will be done in about an hour. (The students are taking a test.)
By the way, my new common core textbook is more consistent with what we learn with modeling than the old one was. I love that they use a “linear system” for students to determine the coefficients of a quadratic equation. (I know it is not related to what we are doing now but I thought it was interesting.)
On Jun 13, 2016, at 10:02 AM, bopangpsy notifications@github.com wrote:
I agree. Then everything for the first presentation looks good now.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D225644029&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=aKRII423ExV8HR9Jj8P1O1Tz1vN_LmEb4ZZKlX_TrGU&s=yNuXk-fYm-E9qpepnC6E3tLBk05fMl1DMASOpZ4Xlhg&e=, or mute the thread https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK066sJc9ONCaix6KWubXXJxUujzzgks5qLY04gaJpZM4Izj99&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=aKRII423ExV8HR9Jj8P1O1Tz1vN_LmEb4ZZKlX_TrGU&s=GI2JJwlplrhS4AWdLBHunJSm5PvC7ZhLi1xRcxSjEGY&e=.
I think we could make the political party predictor variable work. We can always add tot he model, play around with it, if we have time.
I'm feeling comfortable about presenting tomorrow. Is there anything that should be added to the presentation?
I think it looks good. Good Luck!
Yes, good luck and thank you.
On Jun 15, 2016, at 5:14 AM, Joseph Blubaugh notifications@github.com wrote:
I think it looks good. Good Luck!
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D226169497&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=fXrGE_nGTJTFxglD_wagkJ07CyGA6VIAOM9CxKhfGHY&s=q0F-3OzE6x4Re6moWJfxsbY_HhCgeNCfah9YRX2gp58&e=, or mute the thread https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK06mGNGcKaUxFiR-2DRfIssWbhi3IDEks5qL-2DywgaJpZM4Izj99&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=fXrGE_nGTJTFxglD_wagkJ07CyGA6VIAOM9CxKhfGHY&s=uSAMetdDRYml9jxiyZdFg-qQFg0QpCTO_geuaIx5DuY&e=.
Looks good. Good luck!
On Wed, Jun 15, 2016 at 8:05 AM, Alison notifications@github.com wrote:
Yes, good luck and thank you.
On Jun 15, 2016, at 5:14 AM, Joseph Blubaugh notifications@github.com
wrote:I think it looks good. Good Luck!
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub <
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D226169497&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=fXrGE_nGTJTFxglD_wagkJ07CyGA6VIAOM9CxKhfGHY&s=q0F-3OzE6x4Re6moWJfxsbY_HhCgeNCfah9YRX2gp58&e=>,
or mute the thread <
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK06mGNGcKaUxFiR-2DRfIssWbhi3IDEks5qL-2DywgaJpZM4Izj99&d=AwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=fXrGE_nGTJTFxglD_wagkJ07CyGA6VIAOM9CxKhfGHY&s=uSAMetdDRYml9jxiyZdFg-qQFg0QpCTO_geuaIx5DuY&e=
.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AKL-emB-8XehxOMF_TPFutVPHX_F8rwJks5qL_irgaJpZM4Izj99
.
Great job on the presentation, Sean! Thanks!
Just finished up yesterdays lecture, well done Sean.
Here are some things we can mention in our next presentation:
I performed the ADF (augmented Dickey-Fuller) test for stationarity. We haven't talked about it in class yet, but it seems to be used pretty commonly. The high p-value suggests that we do not have a stationary model with just the raw unemployment data. This is confirmed with the non-constant variance and time-dependent mean that we observed in our exploratory analysis.
Augmented Dickey-Fuller Test
Dickey-Fuller = -1.4266, Lag order = 6, p-value = 0.8176
alternative hypothesis: stationary
I then plotted the first difference, and it appears relatively stationary.
This is confirmed by the ADF test on the first differenced data, which has a very small p-value:
Augmented Dickey-Fuller Test
Dickey-Fuller = -4.3501, Lag order = 6, p-value = 0.01
alternative hypothesis: stationary
I tried some basic models on the data after taking the first difference. Out of all the models I tried, ARMA p = q = 2 appeared to have the lowest AICc. Here is the code I used.
Here is the header of the data frame I used:
date unem idi new hpi spend sales
[1,] 8401 8.0 66.6681 224516 104.40 31283 137020
[2,] 8432 7.8 67.6334 245826 104.08 30264 134462
[3,] 8460 7.4 67.6601 260569 104.68 33794 153025
[4,] 8491 6.9 66.6265 243538 105.57 37257 158615
[5,] 8521 6.8 66.6590 241420 105.91 40124 163519
[6,] 8552 7.2 68.6221 264103 106.70 43842 162964
Look at acf and pacf of differenced data
acf(data[,2], main = "Unemployment")
pacf(data[,2], main = "Unemployment")
Try ARIMA p = 1, q = 0, on differenced data
fit1 <- sarima(diff(data[,2]), 1, 0, 0)
Try ARIMA p = 0, q = 1, on differenced data
fit2 <- sarima(diff(data[,2]), 0, 0, 1)
Try ARIMA p = 1, q = 1, on differenced data
fit3 <- sarima(diff(data[,2]), 1, 0, 1)
Try ARIMA p = 2, q = 1, on differenced data
fit4 <- sarima(diff(data[,2]), 2, 0 ,1)
Try ARIMA p = 1, q = 2, on differenced data
fit5 <- sarima(diff(data[,2]), 1, 0 , 2)
Try ARIMA p = q = 2, on differenced data
fit6 <- sarima(diff(data[,2]), 2, 0, 2)
Display AICc of all modelts
fit1$AICc
fit2$AICc
fit3$AICc
fit4$AICc
fit5$AICc
fit6$AICc
Seems that ARMA with p = q = 2 is best
Here is the plot that sarima( ) automatically generated for the ARMA p = q = 2 model:
I think that looks like a good starting point... my only thought is that the data is still seasonal at this point and that may be tricking the ADF test. One clue is that it looks like lag 12 and lag 24 on your residuals both have high correlation with time t. Even though your plot (top left) looks stationary, i think its because of the seasonality. Here are some plots of the seasonally adjusted unemployment rate of different orders... we may have to go 2 or 3 deep to get stationary data.
Also, in the data folder of the project I have an .rda file that you can load as is that has the econ data already prepped (i included the seasonally adjusted unemployment rate as a column as well). I think this is cleaner than everyone loading the data directly from spreadsheets on their own computers... if you have an Rstudio project or your working directory starts in the STAT626_PROJECT folder, then you should only have to do this to load the dataset.
load("Data/Data_Prep.rda")I have updated your script to do this.
Great idea, thank you.
On Sun, Jun 26, 2016 at 12:44 PM, Joseph Blubaugh notifications@github.com
wrote:
I think that looks like a good starting point... my only thought is that
the data is still seasonal at this point and that may be tricking the ADF
test. One clue is that it looks like lag 12 and lag 24 on your residuals
both have high correlation with time t. Even though your plot (top left)
looks stationary, i think its because of the seasonality. Here are some
plots of the seasonally adjusted unemployment rate of different orders...
we may have to go 2 or 3 deep to get stationary data.[image: unemployment_differencing_plots]
https://urldefense.proofpoint.com/v2/url?u=https-3A__cloud.githubusercontent.com_assets_3339909_16364349_a77c74e4-2D3baa-2D11e6-2D9644-2Dca48f6be3bf1.png&d=CwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=Uqb2t_q1xLHYoU5aukJl9GZ2ErgnWICuQUkMdtlllHA&s=MTxe-XVS0Kw0lUyJ2N08NMLqcS4OQdB3pr-yw9ceZaQ&e=Also, in the data folder of the project I have an .rda file that you can
load as is that has the econ data already prepped (i included the
seasonally adjusted unemployment rate as a column as well). I think this is
cleaner than everyone loading the data directly from spreadsheets on their
own computers... if you have an Rstudio project or your working directory
starts in the STAT626_PROJECT folder, then you should only have to do this
to load the dataset.load("Data/Data_Prep.rda")
I have updated your script to do this.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D228618649&d=CwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=Uqb2t_q1xLHYoU5aukJl9GZ2ErgnWICuQUkMdtlllHA&s=IOkb4e-YTp6FrVCJxQLw9vsOvjmCqPigFdsK7gGu25Q&e=,
or mute the thread
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK0-2DHKcPVGsXfsfgmR-5F-5FskINSXEf55ks5qPtaKgaJpZM4Izj99&d=CwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=Uqb2t_q1xLHYoU5aukJl9GZ2ErgnWICuQUkMdtlllHA&s=W6mWw7hyraAavowCYEoTrQQnmlJIWJS8Wz5i2wxUpdo&e=
.
Does anyone know the expectations for the next presentation?
Actually I want to post a correction.. my plot was showing the trend data differences and not the seasonally adjusted data... here is the correct plot. It looks like the 2nd order difference definitely makes unemployment stationary. I have updated the data prep code and data_prep.rda file as well.
Yes, 2nd order on the seasonally adjusted looks good. I will start looking at some preliminary models.
We can basically follow the steps of Example 3.46 in the text.
I took the second difference (d = 2), as Joseph suggested, then the first seasonal difference (D = 1) with s = 12 (this is common for monthly economic data), as the book did. Below is the plot of the transformed graphed. It looks pretty stationary (not perfect, but adequate), and we can confirm this with the ADF test (it's cited in other time series texts, but I haven't seen it in ours yet).
After that, the book suggests that you examine the ACF and PACF plots.
First, the book says to look at the seasonal changes in ACF and PACF (h = 12, 24, 36, ...). These seem to indicate that the ACF trails off, and the PACF cuts off after one year (h = 12). This suggests that we let P = 1 and Q = 0.
Next, the book says to look at the ACF and PACF within only the first season (h = 1, 2, ..., 12). The PACF declines slowly, but the ACF cuts off after 1, suggesting we let p = 0, and q = 1.
When we put this all together, we get a (S)ARIMA(0, 2, 1) x (1, 1, 0) with s = 12 model. I fit that model, and here are the diagnostic plots:
And here are the parameter estimates:
Coefficients:
ma1 sar1 constant
-0.8322 -0.4868 0.0348
s.e. 0.0355 0.0536 6.7411
sigma^2 estimated as 0.03703: log likelihood = 54.87, aic = -101.75
Overall, I think the diagnostics look good. The standardized residual plot isn't great, but it isn't terrible. The normal Q-Q and ACF of residuals look pretty good. The p-values for Ljung-Box are not amazing, but they are at least above the line until H = 20, which I believe the book says is a decent cutoff.
Also, here is the mathematical representation of our model, which we will probably need at some point.
Please check everything I did. I am sure I messed up somewhere. I will post my R code and plots so everyone can check. Also, we should come up with a couple of other models to test, so if you interpret the ACF and PACF differently, that is great. I am not intending this to be the final model, just a starting point. Hopefully, some of this can go toward next week's presentation.
I am happy to add things to Overleaf, but I don't know how, and I am only familiar with basic Latex in the context of Word's Equation Editor.
Pardon my absence; I haven't done a good job of keeping up with this class lately, but I'm back.
My issue is actually figuring out which model is best. Really, my first instinct is to check the residual plots, but I feel like there are other ways to check this.
Same goes for me. I am looking into a few models. Hopefully will knock a
few out in a couple of days and get back.
On Jun 30, 2016 17:05, "Sean Roberson" notifications@github.com wrote:
Pardon my absence; I haven't done a good job of keeping up with this class
lately, but I'm back.My issue is actually figuring out which model is best. Really, my first
instinct is to check the residual plots, but I feel like there are other
ways to check this.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AS9ZZ-OWRb0iS8HSNUEwVkQEThvTuyNkks5qRD2sgaJpZM4Izj99
.
I am playing around with the models too.
On Jun 30, 2016, at 3:07 PM, pakarshan notifications@github.com wrote:
Same goes for me. I am looking into a few models. Hopefully will knock a
few out in a couple of days and get back.
On Jun 30, 2016 17:05, "Sean Roberson" notifications@github.com wrote:Pardon my absence; I haven't done a good job of keeping up with this class
lately, but I'm back.My issue is actually figuring out which model is best. Really, my first
instinct is to check the residual plots, but I feel like there are other
ways to check this.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AS9ZZ-OWRb0iS8HSNUEwVkQEThvTuyNkks5qRD2sgaJpZM4Izj99
.—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D229802774&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=IG-l9_zFhj9CDD8VOwXk4wPN82DyUGcXTU7JkBNe9Ik&s=qXfS_LJNJnyUfwpPyxM4dwNfkhvnbXf0474zLtpWRn8&e=, or mute the thread https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK09O7lv9bttozrPzXc9WEaN6gNVYuks5qRD4RgaJpZM4Izj99&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=IG-l9_zFhj9CDD8VOwXk4wPN82DyUGcXTU7JkBNe9Ik&s=vOgK6_aoLo0ywBSkTroLeyP7GqfUlWhUZzHWBIROI8Q&e=.
I think we need another thread for posting proposed models so we can keep this discussion related to the presentation material... i have another thread started specifically for modeling discussions. I have copied Travis' post to that thread to get us started. I would suggest everyone post similar findings on your proposed models as well as the r code in the /RScrpts folder so we can easily compile and compare the models in a table at the end.
Anyone know if we have class tomorrow? I can have the presentation/write-up done by then if I need to but it would be better if I can work on it some more before sharing.
There will be a video posted. Based on how today went I would say I'm
presenting Monday.
I need to look at the current models and plots to see what would be of
interest to discuss.
On Jul 6, 2016 6:16 PM, "Alison" notifications@github.com wrote:
Anyone know if we have class tomorrow? I can have the presentation/write-up
done by then if I need to but it would be better if I can work on it some
more before sharing.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AS8hey6eSGz9x9ipslnC4beMWeFTxWDRks5qTDdngaJpZM4Izj99
.
Sean, does that mean you are OK with me uploading the presentation tomorrow
night? If you need it sooner, please let me know.
On Wed, Jul 6, 2016 at 4:18 PM, Sean Roberson notifications@github.com
wrote:
There will be a video posted. Based on how today went I would say I'm
presenting Monday.I need to look at the current models and plots to see what would be of
interest to discuss.
On Jul 6, 2016 6:16 PM, "Alison" notifications@github.com wrote:Anyone know if we have class tomorrow? I can have the presentation/write-up
done by then if I need to but it would be better if I can work on it some
more before sharing.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<
#5 (comment),
or mute the thread
<
https://github.com/notifications/unsubscribe/AS8hey6eSGz9x9ipslnC4beMWeFTxWDRks5qTDdngaJpZM4Izj99.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D230936062&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=pBHhbQyJCkTqXWZpCk__QqiS36b_eIXgLKXC_75SHcc&s=95VAv3NwK2ypO_WeEJXQSLD3eqx3i5-scGUOCggkCW0&e=,
or mute the thread
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK0zh4cn9gC0-5F1-2DeeOjC9Lp3q6FHTUks5qTDfLgaJpZM4Izj99&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=pBHhbQyJCkTqXWZpCk__QqiS36b_eIXgLKXC_75SHcc&s=50kWxKliRnWaELtjQbqKY7aOMEJExuVMgP6ywoEtnqU&e=
.
I'm okay with that.
On Jul 6, 2016 11:50 PM, "Alison" notifications@github.com wrote:
Sean, does that mean you are OK with me uploading the presentation tomorrow
night? If you need it sooner, please let me know.On Wed, Jul 6, 2016 at 4:18 PM, Sean Roberson notifications@github.com
wrote:There will be a video posted. Based on how today went I would say I'm
presenting Monday.I need to look at the current models and plots to see what would be of
interest to discuss.
On Jul 6, 2016 6:16 PM, "Alison" notifications@github.com wrote:Anyone know if we have class tomorrow? I can have the
presentation/write-up
done by then if I need to but it would be better if I can work on it some
more before sharing.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<,
or mute the thread
<https://github.com/notifications/unsubscribe/AS8hey6eSGz9x9ipslnC4beMWeFTxWDRks5qTDdngaJpZM4Izj99
.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
<
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D230936062&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=pBHhbQyJCkTqXWZpCk__QqiS36b_eIXgLKXC_75SHcc&s=95VAv3NwK2ypO_WeEJXQSLD3eqx3i5-scGUOCggkCW0&e=
,
or mute the thread
<
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK0zh4cn9gC0-5F1-2DeeOjC9Lp3q6FHTUks5qTDfLgaJpZM4Izj99&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=pBHhbQyJCkTqXWZpCk__QqiS36b_eIXgLKXC_75SHcc&s=50kWxKliRnWaELtjQbqKY7aOMEJExuVMgP6ywoEtnqU&e=.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AS8hez8Yqj7BPPPKLahW6dBwGtkPQA_Tks5qTIWbgaJpZM4Izj99
.
Here is an inital template for a presentation. Please let me know about any changes you'd like made.
I will add the commentary for the write-up.
https://www.overleaf.com/5646560qcxtqg
Here is a newer version: I like this layout better: https://www.overleaf.com/5654811dmsqbs
Okay, I just presented. You can probably hear some shakiness because, in all honesty, I didn't really take detailed markups in my notes for this presentation. I still think I did an okay job of showing the class our models.
As far as the Ljung-Box statistics and explaining them... I think I totally messed that one up.
I just listened to the lecture. Its was hard to hear what the audience was
asking, but I thought you did a good job. If the questions were asking why
the regressors didnt help the model fit I think its because we were not
using them properly. We can get that cleaned up before the next
presentation. Good work.
On Mon, Jul 11, 2016 at 3:42 PM, Sean Roberson notifications@github.com
wrote:
Okay, I just presented. You can probably hear some shakiness because, in
all honesty, I didn't really take detailed markups in my notes for this
presentation. I still think I did an okay job of showing the class our
models.As far as the Ljung-Box statistics and explaining them... I think I
totally messed that one up.—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/ADL2hUxG4OdEUrMdjb5hHqTZBSOhbrzuks5qUqrKgaJpZM4Izj99
.
Good job Sean. I will get the write-up finished after work tomorrow so we
can turn it in.
On Mon, Jul 11, 2016 at 4:32 PM, Joseph Blubaugh notifications@github.com
wrote:
I just listened to the lecture. Its was hard to hear what the audience was
asking, but I thought you did a good job. If the questions were asking why
the regressors didnt help the model fit I think its because we were not
using them properly. We can get that cleaned up before the next
presentation. Good work.On Mon, Jul 11, 2016 at 3:42 PM, Sean Roberson notifications@github.com
wrote:Okay, I just presented. You can probably hear some shakiness because, in
all honesty, I didn't really take detailed markups in my notes for this
presentation. I still think I did an okay job of showing the class our
models.As far as the Ljung-Box statistics and explaining them... I think I
totally messed that one up.—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<
#5 (comment)
,
or mute the thread
<
https://github.com/notifications/unsubscribe/ADL2hUxG4OdEUrMdjb5hHqTZBSOhbrzuks5qUqrKgaJpZM4Izj99.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D231896287&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=ZUB64c3ot3FXKT7DLTFxNH5ZtiMwTjG6E8uVPtbz5sg&s=L3MsRqyK1Ze0lZnhO6tiBenkKFQ-5MC3Pwl36PlAZgI&e=,
or mute the thread
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK0zXZ4vIITdZaolbcTW6MYpaF387Tks5qUtJ-2DgaJpZM4Izj99&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=ZUB64c3ot3FXKT7DLTFxNH5ZtiMwTjG6E8uVPtbz5sg&s=KXR_vBuZxNczlJQPF2KJ72f4BXwsOeqDyGhgEdphEHg&e=
.
The presentation was good. Good job, Sean.
On Mon, Jul 11, 2016 at 9:35 PM, Alison notifications@github.com wrote:
Good job Sean. I will get the write-up finished after work tomorrow so we
can turn it in.On Mon, Jul 11, 2016 at 4:32 PM, Joseph Blubaugh <notifications@github.com
wrote:
I just listened to the lecture. Its was hard to hear what the audience
was
asking, but I thought you did a good job. If the questions were asking
why
the regressors didnt help the model fit I think its because we were not
using them properly. We can get that cleaned up before the next
presentation. Good work.On Mon, Jul 11, 2016 at 3:42 PM, Sean Roberson <notifications@github.com
wrote:
Okay, I just presented. You can probably hear some shakiness because,
in
all honesty, I didn't really take detailed markups in my notes for this
presentation. I still think I did an okay job of showing the class our
models.As far as the Ljung-Box statistics and explaining them... I think I
totally messed that one up.—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<,
or mute the thread
<https://github.com/notifications/unsubscribe/ADL2hUxG4OdEUrMdjb5hHqTZBSOhbrzuks5qUqrKgaJpZM4Izj99
.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
<
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D231896287&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=ZUB64c3ot3FXKT7DLTFxNH5ZtiMwTjG6E8uVPtbz5sg&s=L3MsRqyK1Ze0lZnhO6tiBenkKFQ-5MC3Pwl36PlAZgI&e=
,
or mute the thread
<
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe_AOQK0zXZ4vIITdZaolbcTW6MYpaF387Tks5qUtJ-2DgaJpZM4Izj99&d=CwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=ZUB64c3ot3FXKT7DLTFxNH5ZtiMwTjG6E8uVPtbz5sg&s=KXR_vBuZxNczlJQPF2KJ72f4BXwsOeqDyGhgEdphEHg&e=.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AKL-egIIJLojmxVBP6p36s_IzpXTarvbks5qUv13gaJpZM4Izj99
.
Based on the refinements made over the last few days, here's a tentative list of talking points I have:
- Recap: shape and nature of the data
- Preliminary models, strengths and weaknesses
- Refinements, inclusion of VAR models
- Forecast comparisons, strength of new models (correlation among predictors?)
- Final model decision, forecast analysis
Feel free to add to this list so I can be on target for Tuesday.
That looks good. I'll start putting together the slides
On Sun, Jul 24, 2016 at 7:41 AM, Sean Roberson notifications@github.com
wrote:
Based on the refinements made over the last few days, here's a tentative
list of talking points I have:
Recap: shape and nature of the data
Preliminary models, strengths and weaknesses
Refinements, inclusion of VAR models
Forecast comparisons, strength of new models (correlation among
predictors?)Final model decision, forecast analysis
Feel free to add to this list so I can be on target for Tuesday.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D234780868&d=CwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=FMFATJCe0JgGnq47BApAEO1CkXXZ2_kok2QgP5tDt6U&s=iBRYsaC6wIp-xlY7nVjk0IZlp8U9C_i2tAzdrsOR9sI&e=,
or mute the thread
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AOQK09fd-2DUg5eLWp3f9pjM6jZ56TTZFDks5qY3l-2DgaJpZM4Izj99&d=CwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=FMFATJCe0JgGnq47BApAEO1CkXXZ2_kok2QgP5tDt6U&s=uP_tNut3_rux4rVhc3URZmWUCdpjTB-gMkZsj4JuYJs&e=
.
I am going to keep working on the final write-up for now. I put what Sean's
outline into slides (I know that these won't be the final section titles).
If anyone wants to start fleshing out ideas on the ppt the link on overleaf
is https://www.overleaf.com/5748439mwhjrx. You can put in notes or images,
if you'd like.
Otherwise, I'll work on the actual slides after the write-up is done.
On Sun, Jul 24, 2016 at 7:48 AM, Alison Shelton ashelton@tamu.edu wrote:
That looks good. I'll start putting together the slides
On Sun, Jul 24, 2016 at 7:41 AM, Sean Roberson notifications@github.com
wrote:Based on the refinements made over the last few days, here's a tentative
list of talking points I have:
Recap: shape and nature of the data
Preliminary models, strengths and weaknesses
Refinements, inclusion of VAR models
Forecast comparisons, strength of new models (correlation among
predictors?)Final model decision, forecast analysis
Feel free to add to this list so I can be on target for Tuesday.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JestonBlu_STAT626-5FPROJECT_issues_5-23issuecomment-2D234780868&d=CwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=FMFATJCe0JgGnq47BApAEO1CkXXZ2_kok2QgP5tDt6U&s=iBRYsaC6wIp-xlY7nVjk0IZlp8U9C_i2tAzdrsOR9sI&e=,
or mute the thread
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AOQK09fd-2DUg5eLWp3f9pjM6jZ56TTZFDks5qY3l-2DgaJpZM4Izj99&d=CwMCaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=dBombbLWrTfMsnz-PMDDwPElw1Pkbz0FWwrCqmhbgJA&m=FMFATJCe0JgGnq47BApAEO1CkXXZ2_kok2QgP5tDt6U&s=uP_tNut3_rux4rVhc3URZmWUCdpjTB-gMkZsj4JuYJs&e=
.
Is the current final writeup the main.pdf file? Looking at it it looks like there is still mention of the seasonal data in there... I thought we were going to drop that and just talk about the seasonally adjusted data.
I'll make sure to fix that
Get Outlook for iOS
On Sun, Jul 24, 2016 at 10:43 AM -0700, "Joseph Blubaugh" notifications@github.com wrote:
Is the current final writeup the main.pdf file? Looking at it it looks like there is still mention of the seasonal data in there... I thought we were going to drop that and just talk about the seasonally adjusted data.
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub, or mute the thread.
OK, I took out the non seasonally adjusted references and I am putting in our discussions here. When I am done, I will look at the presentation.