About somebody

somebody · ‎11-10-2019

Thanks. This solves half of my problem. Is there a way to make the 2 lines with the same color to have a solid and a dash pattern? I tried datalinepatterns=(solid dash) datacontrastcolors=(black red) but SAS cycles the colour first and then the pattern which is quite frustrating

somebody · ‎11-10-2019

I can do transparency. However, choosing colour with patterns is troublesome as (I think) SAS cycles colours first and then the patterns

somebody · ‎11-09-2019

How can I control the color of the band when I use GROUP option? For example, I have 4 groups and would like to have specific colour for each group, or at least something related to the line color for that group. My code is: proc sgplot data=sp_in_out_stats; styleattrs datalinepatterns=(solid dash) datacontrastcolors=(orange black); band x=t lower=cum_r_excess_lclm upper=cum_r_excess_uclm/group=in_out ; series x=t y=cum_r_excess_mean/group=in_out curvelabel curvelabelloc=outside ; run; Note that I have removed other function/option within the PROC SGPLOT for ease of following. So can you see that in the following graph, each band has a different color. One outcome I would like to achieve is to have 1 color for the top 2 groups (IN_0 IN_1) and another color for the bottom two. Each group would have different levels of transparency.

somebody · ‎11-04-2019

I am not sure what I am missing. I know that using the custom trading days and weekday give different results. My questions is when I use the custom trading days, it takes a long time to create new dates.

somebody · ‎11-04-2019

Here are the datasets. Data_in the input data_wanted is the desired output. tradingdays contains trading days. Take 2014-05-31 for example. its a public holiday so it is not a trading day. If I use 'weekday', SAS will return this day in my output which is not what I want. Thus, I have a dataset that contains valid trading days. Thanks

somebody · ‎11-04-2019

I am doing an event study. So I would like to get 50 trading days before and 50 trading days after the event day. I have a dataset that contains trading dates because I cannot simply use 'weekday' for the INTNX function in SAS. My code is : options intervalds=(tradingdays=tradingdays) ; data data_window(rename=(date=date_t0 date2=date)); set data_in; by stock date; if first.date; format date2 YYMMDD10.; do t=-50 to 50; date2=intnx('tradingdays',date,t) ;output; end; run; Data_in contains 2 variables: stock and date which is the event date. This runs very slow, as compared to using intnx('weekday',date,t). Is there a different way to improve the speed?

somebody · ‎10-23-2019

I did try it. but could not get it to work. I tried: ods pdf file = "graphs.pdf" startpage=never; ods layout gridded columns=2; proc sgplot data=prices; by stock; series x=date y=price; series x=date y= volume/y2axis; run; ods layout end; ods pdf close; I got 4 graphs per page which is nice but they are all in 1 columns. I would like to get 2x2 or 3x3 graphs. That is 3 columns and 3 rows.

somebody · ‎10-23-2019

Hi all. I am trying to print 4 graphs per page to a PDF file. I have many graphs (200 for now) as I am drawing using PROC SGPLOT and BY statement. I searched online for the solution but all of them involve manually indicating the position of each graphs which is not ideal in my case. Is there a solution where I can tell SAS to print 4 graphs into 1 page instead of 1 big graph per page? My code is: proc sgplot data=prices; by stock; series x=date y=price; series x=date y= volume/y2axis; run;

somebody · ‎10-13-2019

I am using PROC EXPAND to compute daily returns. Theoretically, If the variable is empty, the return should be empty (or at least 0). However, PROC EXPAND gives value for the return of empty observations. Why is it? and How to fix it? Here is my code: proc expand data=data_in out=data_out(drop=time); by fund_ticker constituent_ticker; convert total_shares_held=etf_growth/ transformout=(ratio 1 -1); run; And here is the result:

somebody · ‎09-15-2019

Adding the option CVP to the libname command seems to resolve the issue. One minor problem is that the new dataset becomes quite large ,about 1.5-2 time the size of the original dataset. This is perhaps because SAS allocates larger buffer/ storage to read in the data.

somebody · ‎09-12-2019

Does this convert the new dataset to the encoding of my system? I had add CVP to my libname statement and it runs so far. I still cannot check if the new dataset is native to my computer because the dataset is very large. The original dataset is 550gb. I am copying it to another library, hoping that SAS will change the encoding when creating a new dataset. However, the new dataset seems to be bigger than the original one (700 gb now). Maybe this is due to the fact that SAS use UTF-8 encoding which uses more bytes? is this true?

somebody · ‎09-12-2019

The original dataset is a SAS dataset. I am downloading the CSV version to manually import. But the dataset is about 500 gb...

somebody · ‎09-12-2019

I have some datasets downloaded from another source (WRDS). When I first work with these datasets, a NOTE message appears and say "the dataset is in a format the is native to another host, or the file encoding does not match the session endcoding. Cross Environment Data Acess will be used ......". For most datasets, there was no further error message for this. However, I have an Error message with my recent dataset. It just does not copy to another library, or let me sort or do anything. I understand that the dataset uses a different encoding ( WLATIN1 to be exact) whereas my session encoding is UTF-8. Wouldn't UTF-8 be able to read in the data? How should I deal with this issue? How can I change the encoding of the dataset once and then work with it later on?

somebody · ‎09-07-2019

How can I add a text or footnote that takes the value of a variable? So it changes depending on the BY group

somebody · ‎09-01-2019

In the "example_results.xlsx" file, column "excel_MEAN" contains means computed by Excel by using the formula AVERAGE(). Column SAS_MEAN contains means computed by SAS. The 2 variables of interest are : r and tvol which are contained in column _NAME_.

Online Status	Offline
Date Last Visited	‎02-17-2022 12:34 AM

Re: Importing a datetime variable with format yyyy-mm-ddThh:mm:ss.ffff...

Re: Importing a datetime variable with format yyyy-mm-ddThh:mm:ss.ffff...

Importing a datetime variable with format yyyy-mm-ddThh:mm:ss.ffffffff...

Re: Reformat columns in a large dataset

Re: Reformat columns in a large dataset

Re: Reformat columns in a large dataset

Reformat columns in a large dataset

Re: Fixed effects and Newey West S.E combined

What are the differences between t-stats from PROC SURVEYREG and PROC ...

How to get Newey West standard errors for large a dataset/regression?

Re: Reformat columns in a large dataset

Re: Are the standard errors from PROC SURVEYREG White SE?

Re: Is it possible to absorb fixed effects when using Proc Surveyreg?

Re: Is it possible to absorb fixed effects when using Proc Surveyreg?

Re: Size of long vs wide dataset

Specify the pattern of groups in PROC SGPLOT

display a macro date variable in date format with %put

Comparing company names: compged vs complev vs spedis

Re: Calculating correlation/covariance for stocks

Re: matrix vectorization

Re: PROC SGPLOT: control BAND colors

Re: PROC SGPLOT: control BAND colors

PROC SGPLOT: control BAND colors

Re: Getting trading days around an event date

Re: Getting trading days around an event date

Getting trading days around an event date

Re: Multiple graphs per page with many graphs

Multiple graphs per page with many graphs

Proc expand returns values for empty observations

Re: Encoding issue with datasets

Re: Encoding issue with datasets

Re: Encoding issue with datasets

Encoding issue with datasets

PROC SGPLOT: how to add a dynamic footnote/text?

Re: Different means between PROC MEANS and EXCEL