bar graph for countries with horizontal lines for region means

[Paula de Souza Leao Spinola]

Hello. I have a cross section of health spending per country (in 7 different regions) for the year of 2018. I created a variable with the average of health spending per region using the egen command.

I would like to have a graph with the ranking of health spending per country as bars and the means of the 7 different regions as horizontal lines. Is it possible to do so?

Many thanks,
Paula

[Maarten Buis]

To answer your question: Yes, it is possible to do so.

To anticipate the likely follow up question: How you do that depends on the exact data structure. If you can give us an example of your data, then we can build example code that is useful to you. See help dataex

[Paula de Souza Leao Spinola]

Many thanks! This command is super helpful.

Please find below the structure of my data:

clear
input str48 country float cat_reg int percap_total float percap_total_reg byte Selection int year
"United States"          5 9839      7380 1 2015
"Andorra"                2 9203  2798.714 0 2015
"Switzerland"            2 7465  2798.714 1 2015
"Norway"                 2 7024  2798.714 1 2015
"Luxembourg"             2 6530  2798.714 1 2015
"Netherlands"            2 5579  2798.714 1 2015
"Sweden"                 2 5550  2798.714 1 2015
"Germany"                2 5532  2798.714 1 2015
"Ireland"                2 5371  2798.714 1 2015
"Austria"                2 5183  2798.714 1 2015
"Denmark"                2 5144  2798.714 1 2015
"Belgium"                2 4939  2798.714 1 2015
"Canada"                 5 4921      7380 1 2015
"France"                 2 4741  2798.714 1 2015
"Australia"              1 4400 1098.6072 1 2015
"Japan"                  1 4286 1098.6072 1 2015
"United Kingdom"         2 4285  2798.714 1 2015
"Iceland"                2 4205  2798.714 1 2015
"Finland"                2 4101  2798.714 1 2015
"Singapore"              1 3657 1098.6072 0 2015
"New Zealand"            1 3648 1098.6072 1 2015
"Malta"                  4 3642 1419.9524 0 2015
"Italy"                  2 3445  2798.714 1 2015
"Spain"                  2 3363  2798.714 1 2015
"Qatar"                  4 3251 1419.9524 0 2015
"Saudi Arabia"           4 3138 1419.9524 0 2015
"South Korea"            1 2835 1098.6072 1 2015
"Cyprus"                 2 2821  2798.714 0 2015
"Slovenia"               2 2806  2798.714 1 2015
"Portugal"               2 2712  2798.714 1 2015
"Kuwait"                 4 2640 1419.9524 0 2015
"Israel"                 4 2560 1419.9524 1 2015
"Taiwan"                 1 2535 1098.6072 0 2015
"Czech Republic"         2 2534  2798.714 1 2015
"United Arab Emirates"   4 2489 1419.9524 0 2015
"Bahrain"                4 2470 1419.9524 0 2015
"Greece"                 2 2352  2798.714 1 2015
"Slovakia"               2 2216  2798.714 1 2015
"Brunei"                 1 2092 1098.6072 0 2015
"Uruguay"                3 2038  948.9063 1 2015
"Hungary"                2 2031  2798.714 1 2015
"Trinidad and Tobago"    3 2024  948.9063 0 2015
"Chile"                  3 1950  948.9063 1 2015
"Estonia"                2 1946  2798.714 1 2015
"Lithuania"              2 1941  2798.714 0 2015
"Maldives"               6 1850   411.375 0 2015
"The Bahamas"            3 1818  948.9063 0 2015
"Poland"                 2 1757  2798.714 1 2015
"Croatia"                2 1736  2798.714 0 2015
"Oman"                   4 1684 1419.9524 0 2015
"Latvia"                 2 1683  2798.714 1 2015
"Bulgaria"               2 1620  2798.714 0 2015
"Panama"                 3 1588  948.9063 1 2015
"Russian Federation"     2 1544  2798.714 0 2015
"Argentina"              3 1457  948.9063 1 2015
"Brazil"                 3 1431  948.9063 1 2015
"Serbia"                 2 1398  2798.714 0 2015
"Costa Rica"             3 1339  948.9063 1 2015
"Barbados"               3 1237  948.9063 0 2015
"Iran"                   4 1232 1419.9524 0 2015
"Belarus"                2 1232  2798.714 0 2015
"Azerbaijan"             2 1221  2798.714 0 2015
"Lebanon"                4 1207 1419.9524 0 2015
"Antigua and Barbuda"    3 1198  948.9063 0 2015
"Turkmenistan"           2 1171  2798.714 0 2015
"Romania"                2 1128  2798.714 0 2015
"South Africa"           7 1109 282.41666 0 2015
"Mauritius"              7 1094 282.41666 0 2015
"Equatorial Guinea"      7 1089 282.41666 0 2015
"Mexico"                 3 1081  948.9063 1 2015
"Bosnia and Herzegovina" 2 1076  2798.714 0 2015
"Malaysia"               1 1072 1098.6072 0 2015
"Namibia"                7 1033 282.41666 0 2015
"Turkey"                 2 1029  2798.714 1 2015
"Ecuador"                3 1028  948.9063 1 2015
"Algeria"                4 1026 1419.9524 0 2015
"Botswana"               7 1019 282.41666 0 2015
"Kazakhstan"             2 1017  2798.714 0 2015
"Suriname"               3  993  948.9063 0 2015
"Montenegro"             2  985  2798.714 0 2015
"Cuba"                   3  977  948.9063 0 2015
"Seychelles"             7  957 282.41666 0 2015
"Dominican Republic"     3  932  948.9063 0 2015
"Macedonia"              2  921  2798.714 0 2015
"Colombia"               3  861  948.9063 1 2015
"Armenia"                2  849  2798.714 0 2015
"Albania"                2  848  2798.714 0 2015
"Georgia"                2  803  2798.714 0 2015
"Tunisia"                4  791 1419.9524 0 2015
"China"                  1  779 1098.6072 0 2015
"Paraguay"               3  738  948.9063 1 2015
"Jordan"                 4  730 1419.9524 0 2015
"Grenada"                3  715  948.9063 0 2015
"Saint Lucia"            3  714  948.9063 0 2015
"Swaziland"              7  693 282.41666 0 2015
"Peru"                   3  683  948.9063 1 2015
"Thailand"               1  614 1098.6072 0 2015
"Dominica"               3  606  948.9063 0 2015
"Marshall Islands"       1  604 1098.6072 0 2015
"Ukraine"                2  598  2798.714 0 2015
end
label values cat_reg cat_reg
label def cat_reg 1 "East Asia and Pacific", modify
label def cat_reg 2 "Europe and Central Asia", modify
label def cat_reg 3 "Latin America and the Caribbean", modify
label def cat_reg 4 "Middle East and North Africa", modify
label def cat_reg 5 "North America", modify
label def cat_reg 6 "South Asia", modify
label def cat_reg 7 "Sub-Saharan Africa", modify

label var cat_reg "classification of country by region"
label var Selection "selection of countries for the graph"
label var percap_total "Total Health Spending per Capita (thousands of 2017 US$, PPP)"
label var percap_total_reg "Total Health Spending per Capita by Region (thousands of 2017 US$, PPP)"

I wish to have a graph with information of health spending for all countries for which the dummy variable named Selection is equal to 0 as well as the means for the regions. I can either have the means of the regions as bars as well (as if regions where other countries) or as horizontal lines.

I did manage to run the graph at the country level (highlighting Brazil) and another graph at the region level (codes below) but I did not manage to combine them with the command twoway. Could someone please help me out with this?

code for the country level graph

separate percap_total, by(country == "Brazil")

graph bar percap_total0 percap_total1 if year==2015 & Selection==1, ///
over(country, label(angle(90) labsize(vsmall)) sort(percap_total) descending) ///
ytitle("") yscale(off) blabel(total, orientation(vertical) size(vsmall)) legend(off) nofill ///
bar(2, bcolor(red))

code for the regional level graph

graph bar (mean) percap_total_reg if year==2015, ///
over(cat_reg, label(angle(90) labsize(vsmall)) sort(percap_total) descending) ///
ytitle("") yscale(off) blabel(total, orientation(vertical) size(vsmall)) legend(off)

[Nick Cox]

To add several different lines I think you'd have to shift the whole graph to twoway as in essence graph bar and graph twoway can't be mixed.

Is it needed any way? I played with your data and grouped by regions, which easily allows an assessment of typical level. This graph could be further improved e.g. by putting the region labels on two lines each, not one.

NB with my default graph scheme (s1color) the colors are too close to green and red. Something else to fix.

clear
input str48 country float cat_reg int percap_total float percap_total_reg byte Selection int year
"United States"          5 9839      7380 1 2015
"Andorra"                2 9203  2798.714 0 2015
"Switzerland"            2 7465  2798.714 1 2015
"Norway"                 2 7024  2798.714 1 2015
"Luxembourg"             2 6530  2798.714 1 2015
"Netherlands"            2 5579  2798.714 1 2015
"Sweden"                 2 5550  2798.714 1 2015
"Germany"                2 5532  2798.714 1 2015
"Ireland"                2 5371  2798.714 1 2015
"Austria"                2 5183  2798.714 1 2015
"Denmark"                2 5144  2798.714 1 2015
"Belgium"                2 4939  2798.714 1 2015
"Canada"                 5 4921      7380 1 2015
"France"                 2 4741  2798.714 1 2015
"Australia"              1 4400 1098.6072 1 2015
"Japan"                  1 4286 1098.6072 1 2015
"United Kingdom"         2 4285  2798.714 1 2015
"Iceland"                2 4205  2798.714 1 2015
"Finland"                2 4101  2798.714 1 2015
"Singapore"              1 3657 1098.6072 0 2015
"New Zealand"            1 3648 1098.6072 1 2015
"Malta"                  4 3642 1419.9524 0 2015
"Italy"                  2 3445  2798.714 1 2015
"Spain"                  2 3363  2798.714 1 2015
"Qatar"                  4 3251 1419.9524 0 2015
"Saudi Arabia"           4 3138 1419.9524 0 2015
"South Korea"            1 2835 1098.6072 1 2015
"Cyprus"                 2 2821  2798.714 0 2015
"Slovenia"               2 2806  2798.714 1 2015
"Portugal"               2 2712  2798.714 1 2015
"Kuwait"                 4 2640 1419.9524 0 2015
"Israel"                 4 2560 1419.9524 1 2015
"Taiwan"                 1 2535 1098.6072 0 2015
"Czech Republic"         2 2534  2798.714 1 2015
"United Arab Emirates"   4 2489 1419.9524 0 2015
"Bahrain"                4 2470 1419.9524 0 2015
"Greece"                 2 2352  2798.714 1 2015
"Slovakia"               2 2216  2798.714 1 2015
"Brunei"                 1 2092 1098.6072 0 2015
"Uruguay"                3 2038  948.9063 1 2015
"Hungary"                2 2031  2798.714 1 2015
"Trinidad and Tobago"    3 2024  948.9063 0 2015
"Chile"                  3 1950  948.9063 1 2015
"Estonia"                2 1946  2798.714 1 2015
"Lithuania"              2 1941  2798.714 0 2015
"Maldives"               6 1850   411.375 0 2015
"The Bahamas"            3 1818  948.9063 0 2015
"Poland"                 2 1757  2798.714 1 2015
"Croatia"                2 1736  2798.714 0 2015
"Oman"                   4 1684 1419.9524 0 2015
"Latvia"                 2 1683  2798.714 1 2015
"Bulgaria"               2 1620  2798.714 0 2015
"Panama"                 3 1588  948.9063 1 2015
"Russian Federation"     2 1544  2798.714 0 2015
"Argentina"              3 1457  948.9063 1 2015
"Brazil"                 3 1431  948.9063 1 2015
"Serbia"                 2 1398  2798.714 0 2015
"Costa Rica"             3 1339  948.9063 1 2015
"Barbados"               3 1237  948.9063 0 2015
"Iran"                   4 1232 1419.9524 0 2015
"Belarus"                2 1232  2798.714 0 2015
"Azerbaijan"             2 1221  2798.714 0 2015
"Lebanon"                4 1207 1419.9524 0 2015
"Antigua and Barbuda"    3 1198  948.9063 0 2015
"Turkmenistan"           2 1171  2798.714 0 2015
"Romania"                2 1128  2798.714 0 2015
"South Africa"           7 1109 282.41666 0 2015
"Mauritius"              7 1094 282.41666 0 2015
"Equatorial Guinea"      7 1089 282.41666 0 2015
"Mexico"                 3 1081  948.9063 1 2015
"Bosnia and Herzegovina" 2 1076  2798.714 0 2015
"Malaysia"               1 1072 1098.6072 0 2015
"Namibia"                7 1033 282.41666 0 2015
"Turkey"                 2 1029  2798.714 1 2015
"Ecuador"                3 1028  948.9063 1 2015
"Algeria"                4 1026 1419.9524 0 2015
"Botswana"               7 1019 282.41666 0 2015
"Kazakhstan"             2 1017  2798.714 0 2015
"Suriname"               3  993  948.9063 0 2015
"Montenegro"             2  985  2798.714 0 2015
"Cuba"                   3  977  948.9063 0 2015
"Seychelles"             7  957 282.41666 0 2015
"Dominican Republic"     3  932  948.9063 0 2015
"Macedonia"              2  921  2798.714 0 2015
"Colombia"               3  861  948.9063 1 2015
"Armenia"                2  849  2798.714 0 2015
"Albania"                2  848  2798.714 0 2015
"Georgia"                2  803  2798.714 0 2015
"Tunisia"                4  791 1419.9524 0 2015
"China"                  1  779 1098.6072 0 2015
"Paraguay"               3  738  948.9063 1 2015
"Jordan"                 4  730 1419.9524 0 2015
"Grenada"                3  715  948.9063 0 2015
"Saint Lucia"            3  714  948.9063 0 2015
"Swaziland"              7  693 282.41666 0 2015
"Peru"                   3  683  948.9063 1 2015
"Thailand"               1  614 1098.6072 0 2015
"Dominica"               3  606  948.9063 0 2015
"Marshall Islands"       1  604 1098.6072 0 2015
"Ukraine"                2  598  2798.714 0 2015
end
label values cat_reg cat_reg
label def cat_reg 1 "East Asia and Pacific", modify
label def cat_reg 2 "Europe and Central Asia", modify
label def cat_reg 3 "Latin America and the Caribbean", modify
label def cat_reg 4 "Middle East and North Africa", modify
label def cat_reg 5 "North America", modify
label def cat_reg 6 "South Asia", modify
label def cat_reg 7 "Sub-Saharan Africa", modify

label var cat_reg "classification of country by region"
label var Selection "selection of countries for the graph"
label var percap_total "Total Health Spending per Capita (thousands of 2017 US$, PPP)"
label var percap_total_reg "Total Health Spending per Capita by Region (thousands of 2017 US$, PPP)"

separate percap_total, by(country == "Brazil")

graph hbar percap_total0 percap_total1 if year==2015 & Selection==1, ///
over(country, label(labsize(vsmall)) sort(percap_total) descending) ///
over(cat_reg, label(labsize(small)) sort(percap_total) descending) ///
ytitle("") yscale(off) blabel(total, size(vsmall)) legend(off) nofill ///
bar(2, bcolor(red)) name(Paula1, replace) ysc(r(0 10500)) aspect(2)

[Paula de Souza Leao Spinola]

Many thanks Nick. Unfortunately this does not solve my problem. I would like to show only some selected countries in the graph bar (this is why I created the dummy variable named Selection). However I would like to show the average of all regions in the world to make it comparable. Those averages are computed in the variable percap_total_reg. Ideally, I would like to have them as bars in a different color as if they were other countries. Is it possible?

[Nick Cox]

Sure. Rewrite your code in terms of twoway bar.

[Paula de Souza Leao Spinola]

I am trying to do so, but I am having some trouble because most of the options which I used above are not allowed for the twoway command (e.g. over(), blabel(), nofill, etc)

How can I have the xlabel as string?
How can I have it all in a descending order?
How can I have the data values in the columns?

clear
input long country_id str48 country float cat_reg int percap_total float percap_total_reg byte Selection int year
102 "United States"          5 9839      7380 1 2015
 74 "Andorra"                2 9203  2798.714 0 2015
 94 "Switzerland"            2 7465  2798.714 1 2015
 90 "Norway"                 2 7024  2798.714 1 2015
 87 "Luxembourg"             2 6530  2798.714 1 2015
 89 "Netherlands"            2 5579  2798.714 1 2015
 93 "Sweden"                 2 5550  2798.714 1 2015
 81 "Germany"                2 5532  2798.714 1 2015
 84 "Ireland"                2 5371  2798.714 1 2015
 75 "Austria"                2 5183  2798.714 1 2015
 78 "Denmark"                2 5144  2798.714 1 2015
 76 "Belgium"                2 4939  2798.714 1 2015
101 "Canada"                 5 4921      7380 1 2015
 80 "France"                 2 4741  2798.714 1 2015
 71 "Australia"              1 4400 1098.6072 1 2015
 67 "Japan"                  1 4286 1098.6072 1 2015
 95 "United Kingdom"         2 4285  2798.714 1 2015
 83 "Iceland"                2 4205  2798.714 1 2015
 79 "Finland"                2 4101  2798.714 1 2015
 69 "Singapore"              1 3657 1098.6072 0 2015
 72 "New Zealand"            1 3648 1098.6072 1 2015
 88 "Malta"                  4 3642 1419.9524 0 2015
 86 "Italy"                  2 3445  2798.714 1 2015
 92 "Spain"                  2 3363  2798.714 1 2015
151 "Qatar"                  4 3251 1419.9524 0 2015
152 "Saudi Arabia"           4 3138 1419.9524 0 2015
 68 "South Korea"            1 2835 1098.6072 1 2015
 77 "Cyprus"                 2 2821  2798.714 0 2015
 55 "Slovenia"               2 2806  2798.714 1 2015
 91 "Portugal"               2 2712  2798.714 1 2015
145 "Kuwait"                 4 2640 1419.9524 0 2015
 85 "Israel"                 4 2560 1419.9524 1 2015
  8 "Taiwan"                 1 2535 1098.6072 0 2015
 47 "Czech Republic"         2 2534  2798.714 1 2015
156 "United Arab Emirates"   4 2489 1419.9524 0 2015
140 "Bahrain"                4 2470 1419.9524 0 2015
 82 "Greece"                 2 2352  2798.714 1 2015
 54 "Slovakia"               2 2216  2798.714 1 2015
 66 "Brunei"                 1 2092 1098.6072 0 2015
 99 "Uruguay"                3 2038  948.9063 1 2015
 48 "Hungary"                2 2031  2798.714 1 2015
119 "Trinidad and Tobago"    3 2024  948.9063 0 2015
 98 "Chile"                  3 1950  948.9063 1 2015
 58 "Estonia"                2 1946  2798.714 1 2015
 60 "Lithuania"              2 1941  2798.714 0 2015
 14 "Maldives"               6 1850   411.375 0 2015
106 "The Bahamas"            3 1818  948.9063 0 2015
 51 "Poland"                 2 1757  2798.714 1 2015
 46 "Croatia"                2 1736  2798.714 0 2015
150 "Oman"                   4 1684 1419.9524 0 2015
 59 "Latvia"                 2 1683  2798.714 1 2015
 45 "Bulgaria"               2 1620  2798.714 0 2015
132 "Panama"                 3 1588  948.9063 1 2015
 62 "Russian Federation"     2 1544  2798.714 0 2015
 97 "Argentina"              3 1457  948.9063 1 2015
135 "Brazil"                 3 1431  948.9063 1 2015
 53 "Serbia"                 2 1398  2798.714 0 2015
126 "Costa Rica"             3 1339  948.9063 1 2015
107 "Barbados"               3 1237  948.9063 0 2015
 57 "Belarus"                2 1232  2798.714 0 2015
142 "Iran"                   4 1232 1419.9524 0 2015
 34 "Azerbaijan"             2 1221  2798.714 0 2015
146 "Lebanon"                4 1207 1419.9524 0 2015
105 "Antigua and Barbuda"    3 1198  948.9063 0 2015
 40 "Turkmenistan"           2 1171  2798.714 0 2015
 52 "Romania"                2 1128  2798.714 0 2015
196 "South Africa"           7 1109 282.41666 0 2015
183 "Mauritius"              7 1094 282.41666 0 2015
172 "Equatorial Guinea"      7 1089 282.41666 0 2015
130 "Mexico"                 3 1081  948.9063 1 2015
 44 "Bosnia and Herzegovina" 2 1076  2798.714 0 2015
 13 "Malaysia"               1 1072 1098.6072 0 2015
195 "Namibia"                7 1033 282.41666 0 2015
155 "Turkey"                 2 1029  2798.714 1 2015
122 "Ecuador"                3 1028  948.9063 1 2015
139 "Algeria"                4 1026 1419.9524 0 2015
193 "Botswana"               7 1019 282.41666 0 2015
 36 "Kazakhstan"             2 1017  2798.714 0 2015
118 "Suriname"               3  993  948.9063 0 2015
 50 "Montenegro"             2  985  2798.714 0 2015
109 "Cuba"                   3  977  948.9063 0 2015
186 "Seychelles"             7  957 282.41666 0 2015
111 "Dominican Republic"     3  932  948.9063 0 2015
 49 "Macedonia"              2  921  2798.714 0 2015
125 "Colombia"               3  861  948.9063 1 2015
 33 "Armenia"                2  849  2798.714 0 2015
 43 "Albania"                2  848  2798.714 0 2015
 35 "Georgia"                2  803  2798.714 0 2015
154 "Tunisia"                4  791 1419.9524 0 2015
  6 "China"                  1  779 1098.6072 0 2015
136 "Paraguay"               3  738  948.9063 1 2015
144 "Jordan"                 4  730 1419.9524 0 2015
112 "Grenada"                3  715  948.9063 0 2015
116 "Saint Lucia"            3  714  948.9063 0 2015
197 "Swaziland"              7  693 282.41666 0 2015
123 "Peru"                   3  683  948.9063 1 2015
 18 "Thailand"               1  614 1098.6072 0 2015
110 "Dominica"               3  606  948.9063 0 2015
 24 "Marshall Islands"       1  604 1098.6072 0 2015
127 "El Salvador"            3  598  948.9063 1 2015
end
label values cat_reg cat_reg
label def cat_reg 1 "East Asia and Pacific", modify
label def cat_reg 2 "Europe and Central Asia", modify
label def cat_reg 3 "Latin America and the Caribbean", modify
label def cat_reg 4 "Middle East and North Africa", modify
label def cat_reg 5 "North America", modify
label def cat_reg 6 "South Asia", modify
label def cat_reg 7 "Sub-Saharan Africa", modify

label var cat_reg "classification of country by region"
label var Selection "selection of countries for the graph"
label var percap_total "Total Health Spending per Capita (thousands of 2017 US$, PPP)"
label var percap_total_reg "Total Health Spending per Capita by Region (thousands of 2017 US$, PPP)"

* highlighting Brazil in the graph
separate percap_total, by(country == "Brazil")

twoway bar percap_total_reg cat_reg if year==2015, ///
ytitle("") yscale(off) legend(off) || ///
bar percap_total0 percap_total1 country_id if year==2015 & Selection==1, ///
ytitle("") yscale(off) legend(off)

All I would like to do is to combine those two graph bars into a single graph, having the columns in descending order

* country level
graph bar percap_total0 percap_total1 if year==2015 & Selection==1, ///
over(country, label(angle(90) labsize(vsmall)) sort(percap_total) descending) ///
ytitle("") yscale(off) blabel(total, orientation(vertical) size(vsmall)) legend(off) nofill ///
bar(2, bcolor(red))

* category level
graph bar (mean) percap_total_reg if year==2015, ///
over(cat_reg, label(angle(90) labsize(vsmall)) sort(percap_total_reg) descending) ///
ytitle("") yscale(off) blabel(total, orientation(vertical) size(vsmall)) legend(off) 

* combining grafs (did not work)
twoway bar percap_total_reg if year==2015, ///
over(cat_reg, label(angle(90) labsize(vsmall)) sort(percap_total) descending) ///
ytitle("") yscale(off) blabel(total, orientation(vertical) size(vsmall)) legend(off) || ///
bar percap_total0 percap_total1 if year==2015 & Selection==1, ///
over(country, label(angle(90) labsize(vsmall)) sort(percap_total_reg) descending) ///
ytitle("") yscale(off) blabel(total, orientation(vertical) size(vsmall)) legend(off) nofill ///
bar(2, bcolor(red))

[Nick Cox]

Some hints within https://www.stata-journal.com/sjpdf....iclenum=gr0034

Sorry, but the goal doesn't interest me enough for me to give further time to this question.

[bibha dhungel]

P.S I found my answer. Thanx