bar chart width of bars

[Anya hewertson]

Hi i created this bar chart using the bar chart option within the data tab in stata and i produced this:

the data for my income predictor is as follows:

Example generated by -dataex-. To install: ssc install dataex
clear
input double p344p
615.33
602.11
1735.59
876
1228.814
713.184
253.15
1956.725
730.41
194.9
741.76
379.2507692307692
1536.57
1269.43
314.394
1343.825
661.7158974358974
655.1025
680.462
536.8
196.142
309.69
562.302
514.62
1119.865
404.152
144.042
512.992
682.44
303.48
540.27
1956.725
332.252
564.6265
734.052
1956.725
194.914
197.912
507.115
507.55
406.56
356.869
696.885
390.3
1004.0153846153846
466.38
1956.725
955.5225
330.424
919.78
888.65
1020.8195
601.584
665.37
86.5
748.27
292.894
960.974
552.072
732.46
660.142
452.25
442.574
495.4175
721.15
74.77
593.96
392.34
481.042
252.35
518.07
387.03
1956.725
991.28
359.732
820.074
154.002
559.52
148.122
810.04
150.192
472.51935897435897
1307.8
363.752
1044.97
180.78
1892.695
291.062
1072
676.71
713.75
772.292
886.3558974358974
1500.95
447.834
305.87
364.06
189.412
1956.725
681.88
end

However as seen on the graph i attached, each bar represents an income of around £40 a week however i want to increase the width of the bar so that in total each bar represents a higher proportion of income- preferably each bar representing £200. How do i do this?
Thank you

[Andrew Musau]

If you want to graph the distribution of income, see

help hist

and the options start() and width()

With graph bar, you will need to create the categories by hand. Here is one approach using your data.

gen category=.
forval i= 200 (200) 2000{
local j= `i'-200
replace category=cond(inrange(p344p, `j', `i'-1), `i', category)
}

label define category 200 "0 to 199" 400 "200 to 399" 600 "400 to 599" ///
800 "600 to 799" 1000 "800 to 999" 1200 "1000 to 1199" 1400 "1200 to 1399" ///
1600 "1400 to 1599" 1800 "1600 to 1799" 2000 "1800 to 1999"
label values category category

gr hbar (percent) p344p, over(category) bar(1, color(none)) scheme(s1color) ///
ytitle("Percent") title("Income distribution in the sample")

Graph:

[Nick Cox]

Here's another way to bin:

gen category = 200 * floor(whatever/200)

For more, see e.g. https://journals.sagepub.com/doi/pdf...867X0400300413 (public)

https://www.stata-journal.com/articl...article=dm0095 (subscription access until 2021)