This project uses python to read in a .csv file, run it through x12a, and write
the output as a .csv file.  The main part of the code is x.py.  

Stata output .csv

There MUST be a column titled "date" and it must be readable by python.  Python
can read a wide variety of date formats, and it's not that hard to change, but
the default is to take dates formatted like Jan-1998.o

x.py reads in the output.csv file.  It also takes an optional second argument
which is the name of the output .csv file that contains the seasonally adjusted
data from running x12.

Stata input .csv

After x.py runs, it creates a date column and for each variable in the Stata
output .csv file, it has a [var]sa variable, which is the seasonally adjusted
series corresponding to that variable.  It creates a .csv file that can be read
back into Stata 

x12 options

If you want to configure options for x12, check the documentation for x12 and
add text to the spc[n] variables in x.py.  These variables form the
non-data-specific parts of the blocks of text that will be spliced with the NSA
data to create the .spc file, which is an input to x12.

Note that x12 can produce a lot of different output.  By default x.py reads the
table "D 11" of the output.  It is possible to get the seasonal factors instead
by changing "D 11" to "D 10", and it is likely possible to get other outputs,
but I've not tested this.

adjust.do 

This .do file, and the .dta file, are an example of how you would use this in
stata.

To use this on your data in stata, start with adjust.do, and make sure htat
your own data exists in the data set to be written to .csv.  It is important to
have a column called 'date' that will be an appropriately formatted date string
when written to .csv.  I don't think this has to be in the first column of the
.csv, but I wouldn't chance it.