Statistics by sub_group

[Giulia Fantini]

I need to produce statistics on :
** The number of patents, green patents, proportion green, per year (total at the end).

NUMBER OF PATENTS I managed this.
bysort filing_year: asdoc tabstat patent_num, stat(N min max p10 p25 p50 p75 p99)

NUMBER of GREEN PATENTS
I have created a dummy variable (0-9) called "scheme_code" which classify the patents into not green (=0) and green(1-9).
How can I tell STATA to produce statistics for green patents (scheme_code between 1 and 9) per year?

PROPORTION GREEN PER YEAR
How can I produce a proportion of green patents per year statistics?

Thanks

[Carlo Lazzaro]

Giulia:
if your query does not match with https://www.statalist.org/forums/help#adviceextras #4, please share what you typed and waht Stata gave you back, along with and extcerpt/example of your dataset (as per FAQ again). Thanks.

[Nick Cox]

egen not_green = total(green == 0), by(year)
egen green = total(inrange(green, 1, 9)), by(year)
tabdisp year, c(green not_green)

I wouldn't call a variable with 10 distinct values a dummy variable. It's categorical, perhaps ordinal.

[Giulia Fantini]

Thanks Carlo for your answer.

Here my varibles:
- Firm: permno
- year: filing_year
- patents: patent_num
- patent classification btw green and no green is a dummy variable: green_d which takes velues btw 1 and 9 if is green and 0 otherwise

Summary stats:

Variable Obs Mean Std. Dev. Min Max

permno 3,160,451 40501.84 28722.36 10001 93436
filing_year 3,160,451 1993.946 22.30416 1834 2022
patent_num 3,160,451 6693380 2634605 1570604 1.15e+07
scheme_code 3,160,451 .2368975 1.172737 0 9
green_d 3,160,451 .0434232 .2038079 0 1


//Number of patents per year here is my code
bysort filing_year: asdoc tabstat patent_num, stat(N min max p10 p25 p50 p75 p99)

//Number of green patents per year
based on my scheme_code I can see that, overall I have:
- 3,023,214 No Green patents
- 137237 Greent patents which are classes (1-9)

What I would like is to determine how many green patents I have per year. I have run the below code which does not return the outcome that I want as it includes also the No-green patents.
egen n_green = total(green_d), by(permno filing_year)
bysort filing_year: asdoc tabstat n_green, stat(N min max p10 p25 p50 p75 p99)


Hope you can help me.
Thanks

[Giulia Fantini]

Thanks a lot Nick. It worked.
Apologies for the wrong terminology used.

[Giulia Fantini]

I had to determine the number of patents number of firms, citations , avrage Xi_nominal,average Xi_real per year (total at the end)

Here what I have done in the do file:

egen n_patents = total(inrange(scheme_code, 0, 9)), by(permno filing_year) // I computed how many patents per company I have per year

egen n_cites = total(cites), by (permno filing_year) // I computed how many citation per company I have per year

bys permno filing_year: egen avg_xi_nominal = mean(xi_nominal) // I have computed the avergae per company per year of xi_nominal

bys permno filing_year: egen avg_xi_real = mean(xi_real) // I have computed the avergae per company per year of xi_nominal

sort permno filing_year

order patent_num permno filing_year n_patents cites n_cites xi_nominal avg_xi_nominal xi_real avg_xi_real

bysort filing_year: asdoc tabstat n_patents n_cites avg_xi_nominal avg_xi_real, stat(N)

Here is where I try to produce the table but the outcome is not correct.
I would like a table that shows the n_patents, citations (n_cites), avrage Xi_nominal, average Xi_real per year per firm.

Can I kindly ask your help to understand what I do wrong?
Thanks

[Nick Cox]

On the face of it, your problem is with asdoc from SSC, which I leave open to people who use it.

[Giulia Fantini]

How can I produce the same outcome without asdoc, please?

[Nick Cox]

Still open to others as far as I am concerned. I don't use asdoc and am not familiar with what it does, so working out how to replicate it is not something I can undertake.

[Giulia Fantini]

Here is the amended code without asdoc.

** The number of patents number of firms, citations , avrage Xi_nominal,average Xi_real per year (total at the end)

bys permno filing_year: egen n_patents_firm = count(n_patents)// I computed how many patents per company I have per year
bys permno filing_year: egen n_citations = sum(cites) // I computed how many citation per company I have per year
bys permno filing_year: egen avg_xi_nominal = mean(xi_nominal) // I have computed the avergae per company per year of xi_nominal
bys permno filing_year: egen avg_xi_real = mean(xi_real) // I have computed the avergae per company per year of xi_nominal

sort permno filing_year

order patent_num permno filing_year n_patents_firm cites n_citations xi_nominal avg_xi_nominal xi_real avg_xi_real cpc scheme_code cites filing_date issue_date issue_year

tabdisp filing_year, c(n_patents_firm n_citations avg_xi_nominal avg_xi_real)

This is a snapshot of what I get as outcome:



filing_ye
ar n_patents_firm n_citations avg_xi_nominal avg_xi_real

1834 1 7 .2017942 1.151761
1835 3 1 .193488 1.174547
1836 1 3 .0463756 .3455161
1837 1 4 .3967184 2.652035
1838 1 0 .2393945 1.358425
1868 1 1 2.22735 6.967609
1883 1 6 .3180774 1.794472
1887 1 1 .1859463 1.055137
1903 1 1 .4522994 1.388189
1904 1 0 .2148499 1.184709
1907 1 0 .6957758 3.836594
1910 1 0 .5150331 2.905621
1913 1 0 .4253995 2.578819
1914 1 1 .1686057 .9512091
1915 1 0 .0090889 .0518758
1916 4 21 .77889 4.557581
1917 2 6 .0846542 .5110795
1918 1 1 .0090889 .0518758
1919 2 8 .2244583 1.268149
1920 1 2 .098946 .5456004