I am describing below: 1) the variablenames 2) the task to program in Stata.

1)     1）Variables:

HS10:                   HS 10 industry code

country_:                    BLS country code

naics6d:                6digit naics industry code

year

month

item_code:            good identifier

itemID:           good identifier, numerical

ch_USD:         monthly ch_USagainstROW

ch_TP:                   monthly ch_tradepartn~W

ch_fed:                  monthly ch_frb_broad

ch_exrate:             choosing one of the above

dlernew_`n’:         chosen exchange rate measure at lag `n’

dlcpinew_`n':        lagged PPI at lag `n’

deltadollarprice:  monthly logprice change of the item, or if at country level, of the country

2)    Program

-  count the number of items per firm in each month, then averagingover the entire time period per firm. Note that firms are identified by thefirst 10 digits of item_code. Note that when averaging over time, you only wantto keep on observation per firm year and month.

-  Run an autoregressive regression of the type in lines 2400ff of theattached code for firms that have respectively an item count c of 1<c<3,3<=c<5, 5<=c<7, c>7 (four regressions).

-  Output a table with a cumulative sum of coefficients and standard errorsfor each grouping.

-  The file that contains the data is gitatemp.dta.

这用do file应该如何编呢？ 请高手指教！谢谢！