Create publication quality regression tables in Stata with asdoc
ฝัง
- เผยแพร่เมื่อ 21 ก.ย. 2024
- asdoc creates publication quality tables from Stata output in MS Word or RTF format. With option nest, it creates a nested regression table, while without option nest, it a detailed table of regression results. asdoc can also create tables from other Stata commands such as summarize, correlate, ttest, tabulate, and many more.
INSTALLATION :
=============
To install the program, type the following in Stata command window:
ssc install asdoc
SUPPORT WEBSITE:
================
Please visit
FinTechProfesso... for more programs and additional details about the asdoc capabilities.
Thank you so much! Your tutorial is extremely helpful!
Thanks for creating the package. On Mac it adds the tables on the same file which makes it really easy to see changes when exploring the data. I've already recommended asdoc to my peers.
Dear Professor, thank you SO much for providing this amazing package and tutorial. You have saved me hours of tedious formatting on MS Word.
A life-saving solution! Thank you very much!
Please tell your colleagues about asdoc and subscribe to this channel for future updates
Thank you so much! You just saved me hours of formatting tables :)
Wow, asdoc is terrific. Thanks for the detailed tutorial.
Fantastic! May God give you more wisdom and knowledge.
I used asdoc for different types of regression tables. I like the flexibility in reporting regression results. It can produce detailed, nested and wide tables and each table has a nice format and lots of options to control the output. Simply the best. I can imagine how the author might have spent hundreds of hours developing it.
Really a high impact addition to the stata software
Genius my friend ...pure genius!
This is truly amazing and helping me so much!! Thank you!!!
thanks for distributing knowledge
You are a life saver!! Thank you!
Sir well done. This is excellent method and way more better than outreg2 or esttab. Exporting results from STATA was one messy job, and you made it quite easy.
This is a sharp edge to Stata additions, wow!! thanks so much prof.
Dr. Yasir Bin Tariq asked about adding additional statistics to the output. The additional statistics can be added using the option stat. This is documented in the help file, if you type :
help asdoc##4Regressions
These details are listed in Section 3.10 of the help file.
3.10 stat(stats from e()) (used only with option nest)
stat() can be used to report additional regression statistics that are stored in macro e(). For example:
stat (rmse, rss)
Please note that each statistic should be separated by the character comma. Some of the most commonly used e() statistics of
regression models are as follows:
-----------------------
Statistics
------------------------
N Number of observations. Reported by asdoc as default
r2 R-squared. Reported by asdoc as default
r2_a Adjusted r-squared
F F-statistics
rmse RMSE
rss Residual sum of squares
ll Log-likelihood
chi2 Chi-square value
--------------------------------------------------------
Thanks, Really Helpful Sir
You are a genius, god bless you sir
Fantastic and professional. The voice looks like famous Attaullah Shah, am I right? If yes, Love and respect from Beijing China. Your fan...
This is amazing. Highly recommend for easy output.
Woll, this is just what I needed! Thank you for the video
Please subscribe and tell your friends about asdoc
great video, very helpful
Thanks so much! It saves me a lot time from using esttab
Thank you so much!
this is awesome especially the updated package on your website
this is awesome! Make my life much easier. Thanks so much
Please subscribe to the channel
Wow, creating regression table from Stata output was never that easy. Thanks
great work.....a most need addition to the stata library
Thank you
Thank you, this video helped a lot
Appreciated. Looking forward for more videos
Dear professor, thanks for the very explicit explanation. I have a little challenge. I am doing multinimial regression and the asdoc command works just fine. however, when I export it to word, the values become totally different. is there some other method for MNL?
thanks
Dear Sumeet Malik, thanks for the suggestion. I shall add an option in the next update to skip asterisks.
can you help me plz..
This is so much better than outreg2. Thank you for this work! Is there a way to not report the significant stars and report the p values instead along with standard errors? Top journals are now discouraging stargazing (asterisks)...
Dear Sumeet Malik, thanks for the suggestion. I shall add an option in the next update to skip asterisks.
thanks for the tutorial. If I want to report confidence intervals alongside coefficients, what would the command look like?
You can do this in asdocx, see this link asdocx.com/examples/regressions/nested/#sideways
thank you this is very helpful. I am wondering how I can specify value labels in a nested table output? It is currently outputting with the number of the value label but not the value name itself. e.g. 1 bn.gender 2. gender instead of 1.bn male 2. female. Grateful for your clarification
Have you tried the label option. If it does not work, then you may try asdocx which has all these missing features fintechprofessor.com/asdocx/
THANK YOU!!!
Dear Professor thanks a lot, could you also explain the text addition please
Thank you so much professor. this is absolutely life-saving. May I ask is it possible to replace standard error in the parenthesis of the nested table with p value?
In response to your request, I have updated asdoc to include p-values in the nested tables. Please see this post for more details fintechprofessor.com/forums/topic/how-to-report-p-values-in-asdoc-nested-table/
Can I use this set of commands with a panel regression in Stata?
Thank you!
For excluding i.year and i.location dummies, you can use the option drop(i.year i.location) option. For more details, see this detailed blog post fintechprofessor.com/2018/08/15/dropping-i-dummies-from-regression-asdoc-word-stata/
Thank you so much for this, saved lots of time for me. One quick question, how to show label name of categorical variables (e.g. INDUSTRY) in your example by using i.INDUSTRY n the regression instead of creating dummies explicitly for each value of INDUSTRY, like what is used in your example.
Any suggestion will be deeply appreciated.
Thanks for the suggestion. I shall work on it. Please cite asdoc in your work. In-text citation
Tables were created using asdoc, a Stata program written by Shah (2018).
Bibliography
Shah, A. (2018). ASDOC: Stata module to create high-quality tables in MS Word from Stata output. Statistical Software Components S458466, Boston College Department of Economics.
Thank you so much! I am using asdoc now... Can you please provide a link to your LinkedIn profile somewhere?
Thank you for this great package. I just tried to use it with "mi estimate" and it did not work. Is there any way to use this package with mi?
Thank you very much it helped a lot! I have one problem. I am estimating the fixed-effects and got Pseudo R^2 equal to .z instead of the normal one. Do you know the reason? How can I fix it?
because there is not Pseudo R square for this model. Check original results.
Thank you sir, you have made a very messy task very comfortable. If I compare it with outreg, outreg2 and estout package then your work is far better. But one question, I think we can not use it for other tests like “xttest3”, “xttxt0”, “xtserial” ,“vif”, “pesaran” , and stationarity tests. How to use sir “asdoc” in those tests? Thanks once again
Dear Syed Tauseef Ali
Thanks for your kinds words. The mention tests generate one line or two lines of results. Therefore, they can be easily appended to the Word document created by asdoc by using the option text(), see the help file for more details. Please note that three is a built-in mechanism for the results of VIF command in asdoc, so this one does not need any further coding.
Hi! Can we use "asdoc" with "teffects" and "eteffects" ? I have already tested but the exported table loses its nicely formatation and the option "nest" seems not to work though. Thank you!
Hi prof. Is it okay if your control variables to be omitted due to collinearity in multiple regression whilst using asdoc?
If you like asdoc, you shall like asdocx even more. asdocx can export to Excel and has tons of other features, see this link for details fintechprofessor.com/asdocx/
hii sir
thanks for developing such a wonderful 'asdoc' package, in the current exercise where did you defined the 'stars' for level of significance, could you reply, please
I used the defaults. i.e when you do not specify stars. However, you can set your own stars using setstars option, you can read about it here fintechprofessor.com/2018/12/19/asdoc-cutomizing-the-regression-output-ms-word-from-stata-confidence-interval-adding-stars-etc/, read point 4 there
Shahzad Khan has asked :I am using three regression with same dependent and controls but with different independent . I don’t know how to make make a table of it. Note.: I can not use all independent in same model" . A provide a breif answer in this blog post fintechprofessor.com/creating-a-nested-regression-table-with-asdoc-in-stata/
Can you share the codes that appear in the notepad file you are preparing a video?
I tried using asdoc , nest command in Hierarchial regression.l get the output in word, but my models are not in one table.Does asdoc work hireg command in stata ?
Agnes : Support for hireg is not yet added to asdoc.
Such a great contribution to the stata world. Please can you share your do file?
Please subscribe to the channel. I shall try to find it and share with you.
Is there asdoc for sem models? Thanks
Thank you professor. may i ask a question? I want use reg's command that is (i.time) (i.location). Then i need to drop time and location dummy in word by using asdoc. but if i type the command, it will occur the error. how do i do it?
For excluding i.year and i.location dummies, you can use the option drop(i.year i.location) option. For more details, see this detailed blog post fintechprofessor.com/2018/08/15/dropping-i-dummies-from-regression-asdoc-word-stata/
Can we use "asdoc" with "ivreg2"??
How to add information/result like adjusted R square, F-Statistics or Chi-2 Statistics, their p-values etc. in the regression tables. By default it is giving number of observations and R-squared(not giving it in fixed-effects) only.
Dear Dr. Yasir Bin Tariq, thanks for your comments. The additional statistics can be added using the option stat. This is documented in the help file, if you type :
help asdoc##4Regressions
These details are listed in Section 3.10 of the help file.
3.10 stat(stats from e()) (used only with option nest)
stat() can be used to report additional regression statistics that are stored in macro e(). For example:
stat (rmse, rss)
Please note that each statistic should be separated by the character comma. Some of the most commonly used e() statistics of
regression models are as follows:
-----------------------
Statistics
------------------------
N Number of observations. Reported by asdoc as default
r2 R-squared. Reported by asdoc as default
r2_a Adjusted r-squared
F F-statistics
rmse RMSE
rss Residual sum of squares
ll Log-likelihood
chi2 Chi-square value
--------------------------------------------------------
Thank you so much.
One more question? Does asdoc support quietly function?
In some commands yes, such as sum reg cor. However, asdoc converts the log output from some commands, in such cases qui cannot be used such as ameans, tab, etc.
Thank you
I am wondering how to report t stat in the parenthesis rather than the standard error which seems default in asdoc.
I think you are asking about nested tables. In the options, add rep(t). For example,
sysuse auto
asdoc reg price mpg, rep(t) replace nest
asdoc reg price mpg length , rep(t) nest
Hi Professor, I have a question about running a regression model for multiple year, how do you get a table that will show the coefficients for each year? For instance, my sample data runs from 1966 to 2019, how do I get a table that will output the coefficient by year?
This is a case of regressions over groups. Here the grouping variable is the year. In this video th-cam.com/video/rkUU7UiygBU/w-d-xo.html, I show group-wise regressions for companies, however, you can replace companies with years. Specifically, try the wide option, that is shown towards the end of the video.
@@AttaullahShah Thank you for your reply Attaullah, I tried to run a regression grouped by year. I checked my variables with the tutorial, my variables are float variables, same as the tutorial's data. I could make a nice table with the tutorial data. However, when I ran with my own sample, the myfile.doc shows a bunch of nonsense.. Any idea what may cause this? Sorry to bother again.
When I open the Myfile.doc, I am looking at these error messages
\cellx2300
\fs20\pard\intbl\ql {Mean dependent var}\cell
\cellx4050
\fs20\pard\intbl\qr {0.501}\cell
\cellx6350
\fs20\pard\intbl\ql {SD dependent var }\cell
\cellx8100
\fs20\pard\intbl\qr {0.576}\cell
ow}
Dear Stat Professor, I am having trouble putting various regression tables together and the error seems to be cluster(locid), when I type this in, it says option cluster() not allowed, I am doing gls regressions. Please let me know if you know how to resolve this error!! thank you so much
Can you please post the code line that generates the given error?
The regression values asdoc displays in Word are different from what I see in the stata output, what am I doing wrong?
Please post the full asdoc command here or email me at attaullah.shah@imsciences.edu.pk
same for me
Hi, it says that the nest option is not valid.. Help please :)
Please provide more details. What was the exact Stata code that you typed and what was the error message shown on the Stata screen.
Hi Professor
Thank you very much for the video.
I would like to ask you about the add (dummy, yes) command because STATA keeps returning an error: invalid 'yes' r(198) and I cannot seem to figure out why. I hope you will be able to help me. Please see my command here:
asdoc probit H1 J7A SIZE FOREIGN1 STATEOWN1 EXPORTER COMPETITORS COMPETITORS_binary MANEXP SUR
VEYt i.a0 i.countrynew, nest drop(i.a0 i.countrynew) add (Industry dummies, yes, Country dummies, yes)
I also tried:
... replace add (Industry dummies, yes, Country dummies, yes)
but my attempt has also been unsuccessful.
Thanks for your help in advance.
Kind Regards
Viktoria