Histogram: asking for suggestions in order to improve graph

[Michael Duarte Goncalves]

Hi everyone,

I construct the graph that is enclosed, as an histogram.
My question is the following:

• Could anyone give me an advice in order to have this graph more readable please?

I am not yet very proficient in stata, so advices and code tips are welcomed.
Thank you very much.

Best,

Michael

[Michael Duarte Goncalves]

The graph obtained.

[Nick Cox]

I would remove the smoother. Zipcodes as I understand it are not random, but I doubt that smoothed density functions tell you more than the histogram. You're using about 50 bins, in which case a table or annotated bar or dot chart identifying the most common rounded codes in order of frequency might be of interest too.

[Michael Duarte Goncalves]

Hi Nick Cox ,

I missed you. It's great to see you answering me again.

Thanks for the interesting suggestion.
I'll see what I can do with these suggestions, and if I get stuck, I'll come back to this wonderful tool called statalist.

Best,

Michael

[Michael Duarte Goncalves]

Hi everyone,

I am stuck. I obtained this ugly graph. Should anyone help me? I want to construct basically what Nick Cox says in #3.

Here is a data example:

* Example generated by -dataex-. For more info, type help dataex
clear
input float id_numerical long sp_zipcode byte sp_zipcode_twodigits str5 tariff_ekon_id long(power_p1 power_p2 power_p3) float obs_by_zipcode
468766 1001 1 "20TD"  3450 3450 . 333
753801 1001 1 "20TD"  3450 3450 . 333
868864 1001 1 "20DHA" 1000    . . 333
538474 1001 1 "20TD"  3300 3300 . 333
117736 1001 1 "20TD"  2800 2800 . 333
end

And here is my initial code:

scatter obs_by_zipcode sp_zipcode_twodigits [w=obs_by_zipcode], msymbol(circle_hollow)

Best,

Michael

[Michael Duarte Goncalves]

Here is the graph.