One trick I like to use with regression models with many terms is to make a global that specifies those variables, so I don’t need to retype them a bunch. I read in the data as a CSV file, generate a new variable that is the weekly average number of crimes within 1000 feet in the historical crime data (see the working paper for more details).
That will be more productive than me commenting on every individual line. Again check out Ben Jann’s awesome website he made an all the great examples. Here is what I did to change my default graph settings. Net set ado "C:\Users\axw161530\Documents\Stata_PlugIns\V15" *let stata know to search for a new location for stata plug insĪdopath + "C:\Users\axw161530\Documents\Stata_PlugIns\V15" Log using "LogitModels.txt", text replace *Set the working directory and plain text log fileĬd "C:\Users\axw161530\Dropbox\Documents\BLOG\Stata_NiceMargins\Analysis" But it is good for others to know what extra packages they need to fully replicate your results.
#Scatter plot with line of best fit stata install#
The code is currently commented out, as I only need to install it once. So I add a location to adopath that is on my machine, and I also do net set ado to that same location.įinally, for here I ssc installed grstyle and palettes. I don’t have write access there, so to install commands I need to set my own place to install them on my local machine. The next part is partly idiosyncratic to my Stata work set up - I call Stata from a centralized install location here at EPPS in UTD. I use set more off so I don’t have to do the annoying this and tell Stata to keep scrolling down. Start Upįirst in my do file, I have a typical start up that sets the working directory and logs the results to a text file. Data and code to follow along can be downloaded from here.
0 kommentar(er)
