StataCorp StataNow 19.5 (rev. 20260415) (x64)

Free Download StataCorp StataNow 19.5 (rev. 20260415) | 4.0 Gb
StataCorp, a leading developer of software for data science, is proud to announce the launch of a major upgrade to its flagship product,Stata.

The new version, Stata 19, provides many substantive advancements-from new statistical features and expanded reporting to interface improvements designed to enhance how users interact with the software. Stata 19 builds on many features previously announced with StataNow, StataCorp's continuous-release software that delivers new features and software enhancements as soon as they are certified ready by the development team. With StataNow, StataCorp regularly delivers new features to end users throughout the year.


1. svyset and svy have new option dofsubpop. When specified, out-of-subpopulation primary sampling units are not counted in the design degrees of freedom.
2. mean, proportion, ratio, and total have new option dofsubpop. When specified with option over(), each subpopulation estimate gets its own degrees of freedom, and out-of-subpopulation observations are not counted in their degrees of freedom. This option has no effect when option over() is not specified.
3. import spss now supports variable labels up to 256 bytes long.
4. Stata's tracking of temporary matrices has been improved. The number of global matrices no longer affects the time Stata takes to remove temporary matrices created within an exiting program.
5. Online help and the search index have been brought up to date for Stata Journal 26(1).
6. collect layout, when working with results from spregress, failed to correctly parse levels of colname with operator e.. This has been fixed.
For example, suppose you collected the results from
. spregress y ..., ... errorlag(W)
where e.y is a coefficient in the second equation, and arranged the collected results into a table with
. collect layout (colname[_cons e.y]) (result[_r_b])
Before this fix, collect layout would report
(level e.y of dimension colname not found)
and fail to show the requested coefficient.
7. graph use, when displaying a graph produced by twoway heatmap with a dataset that differs from what is currently in memory, displayed the saved graph incorrectly even though the saved .gph file was correct. This has been fixed.
8. (StataNow) import parquet has the following fixes:
a. import parquet for binary columns did not go beyond the first \0. This has been fixed.
b. import parquet, when either option favormemory was specified or a column list was used and when the parquet file contained a column or columns with nested parquet data types (list, struct, or map), produced an error message or imported the wrong columns. This has been fixed.
9. (StataNow) lateffects with method ipwra with binary or fractional outcome models produced standard errors for the LATE parameter that were conservative. For count outcomes, the standard errors were too small. This has been fixed.
10. me commands mecloglog, meglm, meintreg, melogit, menbreg, meologit, meoprobit, mepoisson, meprobit, mestreg, and metobit, when levelvar was specified as _all with one or more spaces before the colon (:), exited with an unhelpful error message. This has been fixed.
11. menl with option rescorrelation(ctar1, t(timevar)) or rescorrelation(ar 1, t(timevar)) incorrectly labeled an estimated parameter as corr that was not a correlation. This parameter is now labeled rho. The corresponding coefficient in e(b) is labeled logitrho or athrho, respectively.
12. ml model, when specified with option svy and when the svyset properties contained a value for dof(#), failed to respect the svyset design degrees of freedom. This has been fixed.
13. Mata function ustrsplit(), when the first element of the string row vector result was an empty string and the row vector contained more than 49 elements, sometimes mistakenly dropped the last element if it was an empty string or added extra empty strings in the middle of the row vector result. This has been fixed.
14. (Unix) In the console version of Stata, the Enter key on a numeric keypad was unrecognized when the terminal sent the VT100 Application Keypad Enter sequence (\eOM). This has been fixed.



Statais the package of choice for panel data, survival analysis, time series, survey data, longitudinal data, multilevel models, SEM, endogenous covariates, sample selection, Bayesian analysis, multiple imputation, LCA, causal inference, and much more. Stata makes modeling and inference easy for all types of outcomes - continuous, binary, ordinal, categorical, count, and fractional. Stata has both a point-and-click interface and a powerful, programmable, and reproducible command syntax.
Stata/MPlets you analyze data in less time on inexpensive multicore laptops and desktops as well as on multiprocessor servers. Use just 2 cores, and your analyses run in one-half to two-thirds the time compared with Stata/SE. Use 4 cores, and cut the time to one-quarter to one-half. Use more for even faster analyses. Stata/MP supports up to 64 cores/processors.
StataNow is Stata with continuous-release version of Stata that offers new features as soon as they are ready. StataNow provides access to new features as soon as they are certified ready by the development team. Those features are added to StataNow via updates as soon as they are available. Thus, StataNow provides access to new features sooner. For instance, StataNow 19 contains features that will also be part of a future major release, Stata 20.

What's it like-Getting started in Stata 19

Want to know how to conduct a basic data analysis using Stata 19? We show you just that. Watch as we show you how to import data into Stata from Excel, how to change the display format of a variable, how to label variables and the values of categorical variables, how to create new variables, how to describe and summarize variables, and how to create several kinds of graphs. Next, we demonstrate how to fit a linear regression model, calculate and graph marginal predictions from the model, and send those commands to Stata's Do-file editor. The video ends with a discussion of how to learn more about Stata using the Help files, PDF documentation, and the search command.
StataCorp LLC, founded in 1985, is a United States based company that offers a comprehensive solution for data science needs. StataCorp LLC's offering, Stata, is a fully integrated software package designed to facilitate data manipulation, visualization, statistics, and reproducible reporting. Its headquarters is located at College Station, Texas, USA.

Owner:StataCorp LLC.
Product Name:StataNow *
Version:19.5 MP (rev. 20260415)
Supported Architectures:x64
Website Home Page :

www.stata.com

Languages Supported:english
System Requirements:Windows & macOs **
Size:4.0 Gb

* Because StataNow is Stata, when we mention "Stata" throughout our website and documentation, we also mean "StataNow". We will be specific about StataNow for features available only in StataNow. And because StataNow is Stata, it is available in all editions (StataNow/MP, StataNow/SE, and StataNow/BE) and on all supported platforms (Windows, Mac, and Linux). Throughout the website and documentation, we will usually refer to just Stata/MP, Stata/SE, and Stata/BE for simplicity.