About Rodcjones

Rodcjones · ‎01-31-2023

@acordes @Rick_SAS @Ksharp @sbxkoenk @StatDave Thank you all for these prompt and thoughtful responses! I've begun investigating each and plan to report back on the results of my learning/experimentation.

Rodcjones · ‎01-26-2023

This is probably more of a statistics question than programming one but hopefully that's ok. I use SAS Enterprise Guide 7.15 and am searching for the best method(s) to use to conduct two hypothesis tests related to the distribution of a categorical variable in a population over time. Using a simulated dataset (named SIM, provided at bottom), the first question I have is: Is this distribution of the categorical variable “stable” over the course of the first 12 time points? Is there a way to statistically answer that question? Though I’m using the same simulated data set for the second question, it can be considered independently from the one above. Let's say in the study from which this data set was drawn, there was a change applied on Jan. 1, 2022 - after 12 time points/halfway through the time series. The question is: In what way(s), if any, is the distribution of the categorical variable different, or changing over time, in 2022? By eyeball, we would maybe guess that the distribution of the four values is stable in the pre-intervention period, but changes over time in the post-intervention period – perhaps A and/or B decrease and C and/or D increase (in terms of proportions of total). For this, I am not enthusiastic about a single chi-square test of homogeneity in which I aggregate 2021 and 2022 and analyze as a 2x4 contingency table. I have it in my head that interrupted time series could yield what we want – but I’m unsure because I’m most interested in being able to detect or describe the change in distribution rather than change in 1 individual categorical variable alone. data SIM; input Month $ CategoricalVar $ Frequency; datalines; 2021-01 A 152 2021-01 B 289 2021-01 C 193 2021-01 D 103 2021-02 A 145 2021-02 B 250 2021-02 C 193 2021-02 D 101 2021-03 A 178 2021-03 B 312 2021-03 C 248 2021-03 D 117 2021-04 A 174 2021-04 B 309 2021-04 C 238 2021-04 D 135 2021-05 A 184 2021-05 B 339 2021-05 C 234 2021-05 D 116 2021-06 A 180 2021-06 B 340 2021-06 C 241 2021-06 D 113 2021-07 A 203 2021-07 B 370 2021-07 C 241 2021-07 D 109 2021-08 A 185 2021-08 B 345 2021-08 C 252 2021-08 D 134 2021-09 A 198 2021-09 B 333 2021-09 C 252 2021-09 D 130 2021-10 A 207 2021-10 B 378 2021-10 C 233 2021-10 D 127 2021-11 A 168 2021-11 B 298 2021-11 C 223 2021-11 D 127 2021-12 A 172 2021-12 B 308 2021-12 C 260 2021-12 D 127 2022-01 A 122 2022-01 B 290 2022-01 C 247 2022-01 D 144 2022-02 A 151 2022-02 B 287 2022-02 C 218 2022-02 D 107 2022-03 A 170 2022-03 B 316 2022-03 C 276 2022-03 D 162 2022-04 A 150 2022-04 B 325 2022-04 C 277 2022-04 D 119 2022-05 A 148 2022-05 B 289 2022-05 C 287 2022-05 D 134 2022-06 A 148 2022-06 B 238 2022-06 C 252 2022-06 D 154 2022-07 A 130 2022-07 B 258 2022-07 C 241 2022-07 D 153 2022-08 A 135 2022-08 B 235 2022-08 C 300 2022-08 D 140 2022-09 A 152 2022-09 B 229 2022-09 C 280 2022-09 D 172 2022-10 A 154 2022-10 B 330 2022-10 C 315 2022-10 D 187 2022-11 A 130 2022-11 B 278 2022-11 C 312 2022-11 D 179 2022-12 A 135 2022-12 B 267 2022-12 C 299 2022-12 D 175 ;

Rodcjones · ‎10-01-2021

Thank you @Watts and @SAS_Rob for the quick and helpful replies!

Rodcjones · ‎10-01-2021

I am struggling to understand the meaning of/difference between the two stats provided for Zelen's Exact Test in the output below. My guess is that the 0.0274 would be used to reject the hypothesis that the crude and adjusted Odds ratios are equal. But the other value 0.0314 gave me pause. Tests for Homogeneity of Odds Ratios ------------------------------------- Breslow-Day-Tarone Chi-Square 5.4384 DF 1 Pr > ChiSq 0.0197 Zelen's Exact Test (P) 0.0274 Exact Pr <= P 0.0314 I reviewed the section on Zelen's test on this page but it didn't clarify it for me. https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.3/statug/statug_freq_details90.htm I'm hoping someone has experience and can help out. Thanks!

Rodcjones · ‎03-17-2021

Simple, elegant and perfect solution - thank you!

Rodcjones · ‎03-17-2021

I am seeking a reference that will help me create what I believe is called a "positive-negative" bar chart (although I'm not sure). Here's an example visual I found on the web at http://beatexcel.com/positive-negative-bar-chart/ . In the SAS documentation I found this reference (specifically the BASELINEINTERCEPT statement) https://documentation.sas.com/?cdcId=pgmsascdc&cdcVersion=9.4_3.5&docsetId=grstatgraph&docsetTarget=n1dlakkx61v72in1k3ebm8rz18qd.htm&locale=en But this is not helpful to me because I prefer/need to do SGPLOT or other SAS Graphics procedures rather than this if possible. This is a sample dataset I'm trying to graph in this way. data have; input day $ value ; datalines; Wednesday 1.13203258 Tuesday 1.055489878 Thursday 0.050301811 Monday -0.145520643 Saturday -0.281669589 Sunday -0.281669589 Friday -1.360262415 ;

Rodcjones · ‎11-24-2020

This is an elegant solution, thank you @mkeintz! For my growth, I was wondering if you have any good references for what is happening in the data _null_ step with the hash stuff. That seems powerful but I don't follow it - and don't recognize the concepts.

Rodcjones · ‎11-23-2020

C doesn't have a sum of 5 because this row 09/21/20 1 Yes C is part of a consecutive run. Once the run has gone 5 days for the same ID, the count should reset, but since this is only the 2nd day of Yes for ID 1, it is associated with the preceding record. I inadvertently seem to have come up with an example and scenario that's tough to follow, so I appreciate the patience and assistance. In this HAVE dataset below I gave a value of 1 to the count_it variable for rows that should be counted. Hopefully that might explain how I came up with the sums in WANT. data have; input date :mmddyy8. id $ event $ location $ count_it; format date MMDDYY8. ; datalines; 09/01/20 1 Yes A 1 09/02/20 1 No A 0 09/03/20 1 Yes A 1 09/04/20 1 No A 0 09/05/20 1 Yes B 1 09/06/20 1 No B 0 09/07/20 1 Yes B 1 09/08/20 1 No B 0 09/09/20 1 Yes B 1 09/10/20 1 Yes B 0 09/11/20 1 Yes C 0 09/12/20 1 No C 0 09/13/20 1 Yes C 1 09/14/20 1 No C 0 09/15/20 1 No C 0 09/16/20 1 Yes C 1 09/17/20 1 No C 0 09/18/20 1 No C 0 09/19/20 1 No C 0 09/20/20 1 Yes C 1 09/21/20 1 Yes C 0 09/22/20 1 No C 0 09/23/20 1 No C 0 09/24/20 1 No C 0 09/25/20 1 No C 0 09/26/20 1 No C 0 09/27/20 1 No C 0 09/28/20 1 Yes C 1 09/29/20 1 No C 0 09/30/20 1 No C 0 10/01/20 1 No C 0 10/02/20 1 No C 0 10/03/20 1 No C 0 10/04/20 1 No C 0 10/05/20 1 Yes Z 1 10/06/20 1 No Z 0 10/07/20 1 No Z 0 10/08/20 1 No Z 0 10/09/20 1 Yes Z 1 10/10/20 1 No Z 0 10/11/20 1 No Z 0 10/12/20 1 No Z 0 10/13/20 1 No Z 0 10/14/20 1 No D 0 09/01/20 2 No A 0 09/02/20 2 No A 0 09/03/20 2 No A 0 09/04/20 2 Yes A 1 09/05/20 2 Yes A 0 09/06/20 2 Yes E 0 09/07/20 2 Yes E 0 09/08/20 2 Yes E 0 09/09/20 2 Yes B 1 09/10/20 2 Yes B 0 09/11/20 2 Yes B 0 09/12/20 2 Yes B 0 09/13/20 2 Yes B 0 09/14/20 2 Yes B 1 09/15/20 2 Yes B 0 09/16/20 2 Yes B 0 09/17/20 2 No B 0 ;

Rodcjones · ‎11-18-2020

The counting rule I needed to implement says to count consecutively-dated observations of a Yes value as 1 until 5 in a row or a No value, then start over. This counting is done at the ID level. In the case of ID 1 it's correct that the No value on 9/2 interrupts the string of Yes's, and counting restarts at the next Yes for ID 1, which is on 9/3.

Rodcjones · ‎11-18-2020

Hi @PeterClemmensen , Location A's total comes from these records: 09/01/20 1 Yes A + 09/03/20 1 Yes A + 09/04/20 2 Yes A 09/05/20 2 Yes A 09/06/20 2 Yes E 09/07/20 2 Yes E 09/08/20 2 Yes E This last group of 5 is counted as 1 (see link to prior post for more on that).

Rodcjones · ‎11-18-2020

This question is a new twist on a previous post that was impressively solved by @AMSAS and @PeterClemmensen at https://communities.sas.com/t5/SAS-Programming/Counting-consecutively-dated-observations-of-Yes-value-as-1/m-p/696575#M212794 The new scenario I'm tackling is how to perform the same counting method by ID but then 1) attribute the count to a new variable and 2) sum by that variable. In the have and want datasets below I seek to count consecutively-dated Yes values, restart the count after 5 ... and use the LOCATION of the first Yes in the run for summing, even when the location changes at some point in the run. data have; input date :mmddyy8. id $ event $ location $; format date MMDDYY8. ; datalines; 09/01/20 1 Yes A 09/02/20 1 No A 09/03/20 1 Yes A 09/04/20 1 No A 09/05/20 1 Yes B 09/06/20 1 No B 09/07/20 1 Yes B 09/08/20 1 No B 09/09/20 1 Yes B 09/10/20 1 Yes B 09/11/20 1 Yes C 09/12/20 1 No C 09/13/20 1 Yes C 09/14/20 1 No C 09/15/20 1 No C 09/16/20 1 Yes C 09/17/20 1 No C 09/18/20 1 No C 09/19/20 1 No C 09/20/20 1 Yes C 09/21/20 1 Yes C 09/22/20 1 No C 09/23/20 1 No C 09/24/20 1 No C 09/25/20 1 No C 09/26/20 1 No C 09/27/20 1 No C 09/28/20 1 Yes C 09/29/20 1 No C 09/30/20 1 No C 10/01/20 1 No C 10/02/20 1 No C 10/03/20 1 No C 10/04/20 1 No C 10/05/20 1 Yes Z 10/06/20 1 No Z 10/07/20 1 No Z 10/08/20 1 No Z 10/09/20 1 Yes Z 10/10/20 1 No Z 10/11/20 1 No Z 10/12/20 1 No Z 10/13/20 1 No Z 10/14/20 1 No D 09/01/20 2 No A 09/02/20 2 No A 09/03/20 2 No A 09/04/20 2 Yes A 09/05/20 2 Yes A 09/06/20 2 Yes E 09/07/20 2 Yes E 09/08/20 2 Yes E 09/09/20 2 Yes B 09/10/20 2 Yes B 09/11/20 2 Yes B 09/12/20 2 Yes B 09/13/20 2 Yes B 09/14/20 2 Yes B 09/15/20 2 Yes B 09/16/20 2 Yes B 09/17/20 2 No B ; run ; data want; input LOCATION $ sum ; datalines; A 3 B 5 C 4 Z 2 ;

Rodcjones · ‎11-10-2020

In applying these great solutions to a project, I was confronted with a new but related challenge. I need a way to prevent the “counter” from resetting until I have two consecutive days of “No”. In other words, in the original post this subset of ID 1, 09/09/20 1 Yes 09/10/20 1 Yes 09/11/20 1 Yes 09/12/20 1 No 09/13/20 1 Yes 09/14/20 1 No 09/15/20 1 No 09/16/20 1 Yes … would be counted as 3 events, the new challenge is for it to be counted as 2 events. The 9/13 Yes needs to be considered associated with the 9/9-9/11 run; the 9/16 Yes is still a new event to be counted because two days of No have occurred. Currently I’m toying with RETAIN as a means of looking back at the previous day’s value but haven’t solved it yet and would welcome the wisdom and skill of others.

Rodcjones · ‎11-05-2020

Thank you @AMSAS and @PeterClemmensen for these excellent and prompt solutions. I tinkered a bit with customizing these to a couple different scenarios/datasets and they work great. I appreciate the links to the extra info on RETAIN as well. I use that from time to time but have a hard time applying it to new problems.

Rodcjones · ‎11-04-2020

I would like to generate sums of records based on a rule that says to count consecutively-dated observations of a Yes value as 1 until 5 in a row or a No value, then start over. Not sure if I explained that well, so I am providing a have data set with raw data, an intermediate data set that shows how I'd like to count*, and the want data set that sums the counts by ID. Here are the data sets: data have; input date :mmddyy8. id $ event $ ; format date MMDDYY8. ; datalines; 09/01/20 1 Yes 09/02/20 1 No 09/03/20 1 Yes 09/04/20 1 No 09/05/20 1 Yes 09/06/20 1 No 09/07/20 1 Yes 09/08/20 1 No 09/09/20 1 Yes 09/10/20 1 Yes 09/11/20 1 Yes 09/12/20 1 No 09/13/20 1 Yes 09/14/20 1 No 09/15/20 1 No 09/16/20 1 Yes 09/17/20 1 No 09/18/20 1 No 09/19/20 1 No 09/20/20 1 Yes 09/21/20 1 Yes 09/22/20 1 No 09/23/20 1 No 09/24/20 1 No 09/25/20 1 No 09/26/20 1 No 09/27/20 1 No 09/28/20 1 Yes 09/29/20 1 No 09/30/20 1 No 10/01/20 1 No 10/02/20 1 No 10/03/20 1 No 10/04/20 1 No 10/05/20 1 Yes 10/06/20 1 No 10/07/20 1 No 10/08/20 1 No 10/09/20 1 Yes 10/10/20 1 No 10/11/20 1 No 10/12/20 1 No 10/13/20 1 No 10/14/20 1 No 09/01/20 2 No 09/02/20 2 No 09/03/20 2 No 09/04/20 2 Yes 09/05/20 2 Yes 09/06/20 2 Yes 09/07/20 2 Yes 09/08/20 2 Yes 09/09/20 2 Yes 09/10/20 2 Yes 09/11/20 2 Yes 09/12/20 2 Yes 09/13/20 2 Yes 09/14/20 2 Yes 09/15/20 2 Yes 09/16/20 2 Yes 09/17/20 2 No ; data intermediate; input date :mmddyy8. id $ event $ count ; format date MMDDYY8. ; datalines; 9/1/2020 1 Yes 1 9/2/2020 1 No 0 9/3/2020 1 Yes 1 9/4/2020 1 No 0 9/5/2020 1 Yes 1 9/6/2020 1 No 0 9/7/2020 1 Yes 1 9/8/2020 1 No 0 9/9/2020 1 Yes 1 9/10/2020 1 Yes 0 9/11/2020 1 Yes 0 9/12/2020 1 No 0 9/13/2020 1 Yes 1 9/14/2020 1 No 0 9/15/2020 1 No 0 9/16/2020 1 Yes 1 9/17/2020 1 No 0 9/18/2020 1 No 0 9/19/2020 1 No 0 9/20/2020 1 Yes 1 9/21/2020 1 Yes 1 9/22/2020 1 No 0 9/23/2020 1 No 0 9/24/2020 1 No 0 9/25/2020 1 No 0 9/26/2020 1 No 0 9/27/2020 1 No 0 9/28/2020 1 Yes 1 9/29/2020 1 No 0 9/30/2020 1 No 0 10/1/2020 1 No 0 10/2/2020 1 No 0 10/3/2020 1 No 0 10/4/2020 1 No 0 10/5/2020 1 Yes 1 10/6/2020 1 No 0 10/7/2020 1 No 0 10/8/2020 1 No 0 10/9/2020 1 Yes 1 10/10/2020 1 No 0 10/11/2020 1 No 0 10/12/2020 1 No 0 10/13/2020 1 No 0 10/14/2020 1 No 0 9/1/2020 2 No 0 9/2/2020 2 No 0 9/3/2020 2 No 0 9/4/2020 2 Yes 1 9/5/2020 2 Yes 0 9/6/2020 2 Yes 0 9/7/2020 2 Yes 0 9/8/2020 2 Yes 0 9/9/2020 2 Yes 1 9/10/2020 2 Yes 0 9/11/2020 2 Yes 0 9/12/2020 2 Yes 0 9/13/2020 2 Yes 0 9/14/2020 2 Yes 1 9/15/2020 2 Yes 0 9/16/2020 2 Yes 0 9/17/2020 2 No 0 ; data want; input id $ sum ; datalines; 1 12 2 3 ; Any ideas would be greatly appreciated! *in the intermediate data set, note especially how these rows are counted. The intermediate data set is intended to illustrate, and is not necessary to generate in the solution if it's not needed. 9/9/2020 1 Yes 1 9/10/2020 1 Yes 0 9/11/2020 1 Yes 0 and 9/4/2020 2 Yes 1 9/5/2020 2 Yes 0 9/6/2020 2 Yes 0 9/7/2020 2 Yes 0 9/8/2020 2 Yes 0 9/9/2020 2 Yes 1 9/10/2020 2 Yes 0 9/11/2020 2 Yes 0 9/12/2020 2 Yes 0 9/13/2020 2 Yes 0 9/14/2020 2 Yes 1 9/15/2020 2 Yes 0 9/16/2020 2 Yes 0

Rodcjones · ‎08-04-2020

Within SAS EG, I am appending data from multiple datasets in WORK to various SQL database tables in the PBIDATA library. I am able to append all data without problems except for one. I get this error NOTE: Appending WORK.SEPSIS to PBIDATA.CEI_SepsisCCG. NOTE: There were 1 observations read from the data set WORK.SEPSIS. NOTE: 0 observations added. NOTE: The data set PBIDATA.CEI_SepsisCCG has . observations and 69 variables. ERROR: CLI execute error: [Microsoft][ODBC SQL Server Driver]Datetime field overflow NOTE: Statements not processed because of errors noted above. NOTE: PROCEDURE APPEND used (Total process time): real time 0.07 seconds cpu time 0.01 seconds NOTE: The SAS System stopped processing this step because of errors. after submitting this code: proc append data=sepsis base=PBIDATA.CEI_SepsisCCG force;run; I have read up on this error and confirmed none of my datetime variables have dates earlier than the year 2015. I also confirmed that all datetime variables in the WORK dataset have a format of DATETIME22.3 and that all datetime variables in the PBIDATA dataset also have a format of DATETIME22.3. One difference -and I'm grasping at straws here - is that some INFORMAT values on the WORK dataset are null (rather than matching the format value of DATETIME22.3). To reitetrate, I have no problems appending other WORK datasets to other PBIDATA datasets, and those tables also include datetime variables. I am wondering if the community has any ideas on what other aspects could be investigated as the cause.

Online Status	Offline
Date Last Visited	‎02-13-2024 01:50 PM

Re: Has the distribution of values in categorical variable been stable...

Has the distribution of values in categorical variable been stable or ...

Re: Interpretation of Zelen's Exact Test for homogeneity of odds ratio...

Interpretation of Zelen's Exact Test for homogeneity of odds ratios

Re: Creating a positive-negative bar chart

Creating a positive-negative bar chart

Re: Counting consecutively-dated observations of a Yes value with attr...

Re: Counting consecutively-dated observations of a Yes value with attr...

Re: Counting consecutively-dated observations of a Yes value with attr...

Re: Counting consecutively-dated observations of a Yes value with attr...

Re: Has the distribution of values in categorical variable been stable...

Re: Has the distribution of values in categorical variable been stable...

Re: Has the distribution of values in categorical variable been stable...

Re: Has the distribution of values in categorical variable been stable...

Re: Has the distribution of values in categorical variable been stable...

Best practices for integrity checks & halting a scheduled EG process w...

Does an oversized SAS EG project log prevent a scheduled task from run...

Re: Has the distribution of values in categorical variable been stable...

Has the distribution of values in categorical variable been stable or ...

Re: Interpretation of Zelen's Exact Test for homogeneity of odds ratio...

Interpretation of Zelen's Exact Test for homogeneity of odds ratios

Re: Creating a positive-negative bar chart

Creating a positive-negative bar chart

Re: Counting consecutively-dated observations of a Yes value with attr...

Re: Counting consecutively-dated observations of a Yes value with attr...

Re: Counting consecutively-dated observations of a Yes value with attr...

Re: Counting consecutively-dated observations of a Yes value with attr...

Counting consecutively-dated observations of a Yes value with attribut...

Re: Counting consecutively-dated observations of Yes value as 1 until ...

Re: Counting consecutively-dated observations of Yes value as 1 until ...

Counting consecutively-dated observations of Yes value as 1 until 5 in...

Troubleshooting "Datetime field overflow" when dates are recent & dt v...