Using the collapse command to create aggregate data from individual-level data using frequency weights. Visit my website for more videos: davidbraudt.com...
David, how do you collapse dummy variables (1/0) or variables with numerical categories (1, 2, 3, 4)? I am assuming you can't take the mean of these variables.
You would probably just ignore the information given by dummy variables unless you actually want to aggregate by those dummy variables. eg by year and gender...
+Thomas Kirkland Hi you can actually create dummies from the a categorical variable ie Frequency of smoking with four options such as 1. Most often. 2. Rarely 3. Never by using tab freq, gen (most_often) command this will generate variables with the name most_often1, 2, 3 where 1=most_often, 2= rarely etc
i have an urgent question: im trying to make a longitudional file from multiple eurobarometer files. i want to measure the changes in issues mentioned as the most important problem over time. can anyone help me. i know so little about stata i dont even know what to look for
Use the merge command on the variable that was used to collapse the data. You will need to use either the m:1 or the 1:m version of merge. For example if you have the collapsed dataset in memory and are merging in the original data (let's assume a case is an individual) it would be a 1:m merge since there is one collapsed observation in the "master" dataset (Stata's language) that corresponds to many observations in the "using" dataset (i.e. the individual-level data you are merging back into the collapsed data). HTH.
Really you are the best teacher i have ever seen thank you very much, you tough me something than i don't understand deeply from my teacher.
Thank you very helpful :)
I found the vedios very informative
David, how do you collapse dummy variables (1/0) or variables with numerical categories (1, 2, 3, 4)? I am assuming you can't take the mean of these variables.
You would probably just ignore the information given by dummy variables unless you actually want to aggregate by those dummy variables. eg by year and gender...
+Thomas Kirkland Hi you can actually create dummies from the a categorical variable ie Frequency of smoking with four options such as 1. Most often. 2. Rarely 3. Never by using tab freq, gen (most_often) command this will generate variables with the name most_often1, 2, 3 where 1=most_often, 2= rarely etc
+Thomas Kirkland +Bosco Kasundu Thanks for replying to this Bosco! I apparently missed it a year back. Hopefully you found an answer back then Thomas.
Thank you
i have an urgent question: im trying to make a longitudional file from multiple eurobarometer files. i want to measure the changes in issues mentioned as the most important problem over time. can anyone help me. i know so little about stata i dont even know what to look for
how do you add the result to the original data set for further analysis?
Use the merge command on the variable that was used to collapse the data. You will need to use either the m:1 or the 1:m version of merge. For example if you have the collapsed dataset in memory and are merging in the original data (let's assume a case is an individual) it would be a 1:m merge since there is one collapsed observation in the "master"
dataset (Stata's language) that corresponds to many observations in the "using" dataset (i.e. the individual-level data you are merging back into the collapsed data). HTH.
David, how can I regress in loop by firm and save the coefficient as a variable in the dataset?
how do we collapse if our variables of interest dummy in nature
Hi, did you figure out how to do this?
thanksss!!!!