Originally posted by William Lisowski
View Post
-
Login or Register
- Log in with
- You are not logged in. You can browse but not post. Login or Register by clicking 'Login or Register' at the top-right of this page. For more information on Statalist, see the FAQ.
-
Thats true, thats why alter I will collapse by (Experiengroup sk_rat). -
With the data in post #1
givesCode:// restrict the calculations to immigrants drop if nacio=="PT" // Clyde's code from post #4 by Expgroup sk_rat_quartile nacio , sort: gen numerator = _N by Expgroup sk_rat_quartile (nacio): gen denominator = _N gen share_percent = 100*numerator/denominator // collapse collapse (first) share=numerator share_percent, by(Expgroup sk_rat_quartile nacio) format %9.1f share_percent list, sepby(Expgroup sk_rat_quartile) abbreviate(20) noobs
Is this what you have in mind?Code:+------------------------------------------------------------+ | Expgroup sk_rat_quartile nacio share share_percent | |------------------------------------------------------------| | 3 3 IR 1 100.0 | |------------------------------------------------------------| | 4 3 GR 1 50.0 | | 4 3 PR 1 50.0 | |------------------------------------------------------------| | 5 3 AO 1 33.3 | | 5 3 SW 1 33.3 | | 5 3 UK 1 33.3 | |------------------------------------------------------------| | 6 3 ES 1 33.3 | | 6 3 SP 1 33.3 | | 6 3 US 1 33.3 | |------------------------------------------------------------| | 7 3 EU 1 25.0 | | 7 3 SP 1 25.0 | | 7 3 UK 2 50.0 | +------------------------------------------------------------+
Comment
-
I really appreciated it.
There are still two problems.
First, I dont get that Collapse part, why using "first"?
Second, now, I am asking you: whats your definition of the sentence "summing up over countries" ? Because I need to sum up over countries in this stage.Comment
-
To select just one observation of any group with the same Expgroup, sk_rat_quartile, and nacioFirst, I dont get that Collapse part, why using "first"?
Consider what the data for Expgroup 7 look like before and after the collapse.
Code:. // before collapse . list if Expgroup==7, abbreviate(20) noobs +------------------------------------------------------------------------------+ | Expgroup sk_rat_quartile nacio numerator denominator share_percent | |------------------------------------------------------------------------------| | 7 3 EU 1 4 25.0 | | 7 3 SP 1 4 25.0 | | 7 3 UK 2 4 50.0 | | 7 3 UK 2 4 50.0 | +------------------------------------------------------------------------------+ . collapse (first) share=numerator share_percent, by(Expgroup sk_rat_quartile nacio) . // after collapse . list if Expgroup==7, abbreviate(20) noobs +------------------------------------------------------------+ | Expgroup sk_rat_quartile nacio share share_percent | |------------------------------------------------------------| | 7 3 EU 1 25.0 | | 7 3 SP 1 25.0 | | 7 3 UK 2 50.0 | +------------------------------------------------------------+
You wrotewhats your definition of the sentence "summing up over countries" ?
So with that in mind, what we have calculated is not what you need. You do not want the percent by Expgroup and sk_rat_quartile of each country, you want the percent by country for each combination of Expgroup and sk_rat_quartile.Specifically, predicted immigrants inflows are going to be calculated by multiplying the total number of newly arriving immigrants from the source country lets call that K at time t (I access to this quantity, no need to compute that, fortunately) by the share of immigrants from source country K that was in skill group (Experience group and skill_ratio) ij in the year 1981.
...
After summing up over countries K, the instrument is constructed as the predicted number of immigrants divided by the total number of workers in a given skill group.
Code:// restrict the calculations to immigrants drop if nacio=="PT" // Clyde's code modified by nacio Expgroup sk_rat_quartile, sort: gen numerator = _N by nacio (Expgroup sk_rat_quartile): gen denominator = _N gen share_percent = 100*numerator/denominator order nacio Expgroup sk_rat_quartile format %9.1f share_percent // before collapse list if nacio=="UK", abbreviate(20) noobs collapse (first) share=numerator share_percent, by(nacio Expgroup sk_rat_quartile) // after collapse list if nacio=="UK", abbreviate(20) noobs
So if there were 300 immigrants from nacio UK, 100 of them would be assigned Expgroup 5 sk_rat_quartile 3, and 200 of them would be assigned Expgroup 7 sk_rat_quartile 3,Code:. // before collapse . list if nacio=="UK", abbreviate(20) noobs +------------------------------------------------------------------------------+ | nacio Expgroup sk_rat_quartile numerator denominator share_percent | |------------------------------------------------------------------------------| | UK 5 3 1 3 33.3 | | UK 7 3 2 3 66.7 | | UK 7 3 2 3 66.7 | +------------------------------------------------------------------------------+ . collapse (first) share=numerator share_percent, by(nacio Expgroup sk_rat_quartile) . // after collapse . list if nacio=="UK", abbreviate(20) noobs +------------------------------------------------------------+ | nacio Expgroup sk_rat_quartile share share_percent | |------------------------------------------------------------| | UK 5 3 1 33.3 | | UK 7 3 2 66.7 | +------------------------------------------------------------+
Added in edit: We would have reached this point much sooner if you had presented, not just a verbal description of what you wanted, but a worked out example using your three observations of UK data. I know you said "I am going to compute the share of immigrants from the source country that are in skill groups (Expgroup & sk_rat_quartile)." but it was not understood to mean "For this data on immigrants giving their source country (nacio) and skill groups (Expgroup & sk_rat_quartile), I want to compute for each country what percentage is in each combination of skill groups (Expgroup & sk_rat_quartile)."
By providing and explaining a worked example you give the reader something to check their understanding against.Last edited by William Lisowski; 11 Feb 2023, 20:23.Comment
-
Prof William,
Many thanks to you.
I am impressed by your unlimited patience. It worked. Thats what I really need.
Your explanation is so clear and I will be careful with providing precise examples and addressing the core problems. Probably, it is due to my not very good English that my verbal explanation misleads you. Sorry about that.
Gratefully yours,
ParisComment
-
Prof, I tried to run this dataset with code #19, and it gives only 1 for all.
what I am going to do is:Code:* Example generated by -dataex-. For more info, type help dataex clear input str7 Regions str24 country int immigrants "Centro" "Spain" 672 " Lisboa" "Spain" 4921 "Centro" "France" 3814 " Lisboa" "France" 1858
Share of Spanish immigrants in Metropolitain= (672+4921)/ (672+ 3814)
Share of French immigrants in metropolitan= 3814+1858) / (3814+672)
I need to split the region, then combine it. sth like this:
So I could be able to compute the share of immigrants onceCode:Region indexregion Country immigrants Centro Lisboa Centro,Lisboa Spain 672+4921 Centro Lisboa Centro,Lisboa Franc 3814+1858
Could you please assist me?
Thank you so much.Last edited by Paris Rira; 13 Feb 2023, 12:57.Comment
-
Code:* Example generated by -dataex-. For more info, type help dataex clear input str7 Regions str24 country int immigrants "Centro" "Spain" 672 " Lisboa" "Spain" 4921 "Centro" "France" 3814 " Lisboa" "France" 1858 end // get rid of all leading and trailing blanks replace Regions = trim(Regions) generate indexregion = Regions replace indexregion = "Centro,Lisboa" if inlist(Regions,"Centro","Lisboa") collapse (first) indexregion (sum) immigrants, by(country) list, clean noobs abbreviate(12)Code:. list, clean noobs abbreviate(12) country indexregion immigrants France Centro,Lisboa 5672 Spain Centro,Lisboa 5593 .Comment
-
It worked perfectly. It is the best as usual. Thanks a lot, Prof William.
Gratefully yours,Code:replace Regions = trim(Regions) generate indexregion = Regions replace indexregion = "Centro,Lisboa" if inlist(Regions,"Centro","Lisboa") collapse (first) indexregion (sum) imm, by(country) list, clean noobs abbreviate(12) egen n_imm= total(imm) g share_imm= imm/n_imm . list +---------------------------------------------------------------------+ | country indexregion imm n_imm share_~m | |---------------------------------------------------------------------| 1. | Angola Centro,Lisboa 11131 58921 .188914 | 2. | Brazil Centro,Lisboa 5009 58921 .0850121 | 3. | Cape Verde Centro,Lisboa 17312 58921 .2938171 | 4. | China Centro,Lisboa 125 58921 .0021215 | 5. | France Centro,Lisboa 5672 58921 .0962645 | |---------------------------------------------------------------------| 6. | Guinea Bissau Centro,Lisboa 918 58921 .0155802 | 7. | India Centro,Lisboa 189 58921 .0032077 | 8. | Italy Centro,Lisboa 525 58921 .0089102 | 9. | Mozambique Centro,Lisboa 2658 58921 .0451113 | 10. | Other African countries Centro,Lisboa 1300 58921 .0220634 | |---------------------------------------------------------------------| 11. | Other American countries Centro,Lisboa 564 58921 .0095721 | 12. | Other Asian countries Centro,Lisboa 863 58921 .0146467 | 13. | Other European countries Centro,Lisboa 4171 58921 .0707897 | 14. | Sao Tome and Principe Centro,Lisboa 1406 58921 .0238625 | 15. | Spain Centro,Lisboa 5593 58921 .0949237 | |---------------------------------------------------------------------| 16. | United Kingdom Centro,Lisboa 1485 58921 .0252032 | +---------------------------------------------------------------------+
ParisComment
-
The code in posts #22 and #23 will not work correctly if the input data contains Regions other than Centro and Lisboa. This code corrects the problem.
Code:* Example generated by -dataex-. For more info, type help dataex clear input str7 Regions str24 country int immigrants "Centro" "Spain" 672 " Lisboa" "Spain" 4921 "Centro" "France" 3814 " Lisboa" "France" 1858 "Oporto" "Spain" 300 "Oporto" "France" 100 end // get rid of all leading and trailing blanks replace Regions = trim(Regions) generate indexregion = Regions replace indexregion = "Centro,Lisboa" if inlist(Regions,"Centro","Lisboa") collapse (sum) immigrants, by(indexregion country) order indexregion country list, clean noobs abbreviate(12) bysort indexregion (country): egen n_imm= total(immigrants) generate share_imm= immigrants/n_imm list, clean noobs abbreviate(12)Code:. list, clean noobs abbreviate(12) indexregion country immigrants Centro,Lisboa France 5672 Centro,Lisboa Spain 5593 Oporto France 100 Oporto Spain 300Code:. list, clean noobs abbreviate(12) indexregion country immigrants n_imm share_imm Centro,Lisboa France 5672 11265 .5035064 Centro,Lisboa Spain 5593 11265 .4964936 Oporto France 100 400 .25 Oporto Spain 300 400 .75Comment
-
Actually, there are only Centro and Lisboa in the Region. But, appreciated for pointing out that.Comment
Comment