Bar Graph the my specific data

[matze pie]

Hallo everyone, I need some help for my Bachelor Thesis, I want to plot with bar graph but I have anytime the same problem " class
member function not found " Before I did the bar graph I have done fisher exact test for one handed hypothesis checking. I need a bar graph which shows the percentage of coming to the walk (Teilnahme) for both groups Persönliche Gruppe(treatment) and Formelle Gruppe(control). The first numeric variable show in which group they are and the second if they come to the walk or not. Someone have an Idea how I can solve this problem ? That could be very helpful for me. Thank you.

Here is my Data

----------------------- copy starting from the next line -----------------------

* Example generated by -dataex-. For more info, type help dataex
clear
input byte ID str3 Mitarbeiter str20 Treatment byte(Dummy Teilnahme)
 1 "K01" "Persönliche Gruppe " 1 0
 2 "H02" "Persönliche Gruppe " 1 0
 3 "G03" "Persönliche Gruppe " 1 0
 4 "D04" "Persönliche Gruppe " 1 0
 5 "H05" "Persönliche Gruppe " 1 0
 6 "M06" "Persönliche Gruppe " 1 0
 7 "A07" "Persönliche Gruppe " 1 0
 8 "L08" "Persönliche Gruppe " 1 0
 9 "N09" "Persönliche Gruppe " 1 0
10 "P10" "Persönliche Gruppe " 1 0
11 "D11" "Persönliche Gruppe " 1 0
12 "K12" "Persönliche Gruppe " 1 0
13 "P13" "Persönliche Gruppe " 1 1
14 "B14" "Persönliche Gruppe " 1 0
15 "H15" "Persönliche Gruppe " 1 0
16 "F16" "Persönliche Gruppe " 1 0
17 "E17" "Persönliche Gruppe " 1 0
18 "B18" "Persönliche Gruppe " 1 0
19 "R19" "Persönliche Gruppe " 1 0
20 "K20" "Persönliche Gruppe " 1 0
21 "W21" "Persönliche Gruppe " 1 0
22 "L22" "Persönliche Gruppe " 1 0
23 "M23" "Persönliche Gruppe " 1 0
24 "T24" "Persönliche Gruppe " 1 0
25 "W25" "Formelle Gruppe"      0 0
26 "W26" "Formelle Gruppe"      0 0
27 "B27" "Formelle Gruppe"      0 0
28 "C28" "Formelle Gruppe"      0 0
29 "M29" "Formelle Gruppe"      0 0
30 "S30" "Formelle Gruppe"      0 0
31 "A31" "Formelle Gruppe"      0 0
32 "D32" "Formelle Gruppe"      0 0
33 "B33" "Formelle Gruppe"      0 0
34 "S34" "Formelle Gruppe"      0 0
35 "O35" "Formelle Gruppe"      0 0
36 "M36" "Formelle Gruppe"      0 0
37 "H37" "Formelle Gruppe"      0 0
38 "G38" "Formelle Gruppe"      0 0
39 "B39" "Formelle Gruppe"      0 1
40 "K40" "Formelle Gruppe"      0 0
41 "S41" "Formelle Gruppe"      0 0
42 "H42" "Formelle Gruppe"      0 0
43 "F43" "Formelle Gruppe"      0 0
44 "W44" "Formelle Gruppe"      0 0
45 "D45" "Formelle Gruppe"      0 0
46 "K46" "Formelle Gruppe"      0 0
47 "M47" "Formelle Gruppe"      0 0
48 "B48" "Formelle Gruppe"      0 0
end

------------------ copy up to and including the previous line ------------------

[Nick Cox]

There is a problem with the code you are trying. To get specific advice on what you did wrong, please show us that code.

You seek answers Yes or No, but the only question here is

Someone have an Idea how I can solve this problem?

to which my reply is that this code may help.

* Example generated by -dataex-. For more info, type help dataex
clear
input byte ID str3 Mitarbeiter str20 Treatment byte(Dummy Teilnahme)
 1 "K01" "Persönliche Gruppe " 1 0
 2 "H02" "Persönliche Gruppe " 1 0
 3 "G03" "Persönliche Gruppe " 1 0
 4 "D04" "Persönliche Gruppe " 1 0
 5 "H05" "Persönliche Gruppe " 1 0
 6 "M06" "Persönliche Gruppe " 1 0
 7 "A07" "Persönliche Gruppe " 1 0
 8 "L08" "Persönliche Gruppe " 1 0
 9 "N09" "Persönliche Gruppe " 1 0
10 "P10" "Persönliche Gruppe " 1 0
11 "D11" "Persönliche Gruppe " 1 0
12 "K12" "Persönliche Gruppe " 1 0
13 "P13" "Persönliche Gruppe " 1 1
14 "B14" "Persönliche Gruppe " 1 0
15 "H15" "Persönliche Gruppe " 1 0
16 "F16" "Persönliche Gruppe " 1 0
17 "E17" "Persönliche Gruppe " 1 0
18 "B18" "Persönliche Gruppe " 1 0
19 "R19" "Persönliche Gruppe " 1 0
20 "K20" "Persönliche Gruppe " 1 0
21 "W21" "Persönliche Gruppe " 1 0
22 "L22" "Persönliche Gruppe " 1 0
23 "M23" "Persönliche Gruppe " 1 0
24 "T24" "Persönliche Gruppe " 1 0
25 "W25" "Formelle Gruppe"      0 0
26 "W26" "Formelle Gruppe"      0 0
27 "B27" "Formelle Gruppe"      0 0
28 "C28" "Formelle Gruppe"      0 0
29 "M29" "Formelle Gruppe"      0 0
30 "S30" "Formelle Gruppe"      0 0
31 "A31" "Formelle Gruppe"      0 0
32 "D32" "Formelle Gruppe"      0 0
33 "B33" "Formelle Gruppe"      0 0
34 "S34" "Formelle Gruppe"      0 0
35 "O35" "Formelle Gruppe"      0 0
36 "M36" "Formelle Gruppe"      0 0
37 "H37" "Formelle Gruppe"      0 0
38 "G38" "Formelle Gruppe"      0 0
39 "B39" "Formelle Gruppe"      0 1
40 "K40" "Formelle Gruppe"      0 0
41 "S41" "Formelle Gruppe"      0 0
42 "H42" "Formelle Gruppe"      0 0
43 "F43" "Formelle Gruppe"      0 0
44 "W44" "Formelle Gruppe"      0 0
45 "D45" "Formelle Gruppe"      0 0
46 "K46" "Formelle Gruppe"      0 0
47 "M47" "Formelle Gruppe"      0 0
48 "B48" "Formelle Gruppe"      0 0
end

graph hbar Teilnahme, over(Treatment) over(Dummy) yla(0 0.01 "1" 0.02 "2" 0.03 "3" 0.04 "4") ytitle(Percent) name(G1, replace)

graph hbar Teilnahme, over(Dummy) over(Treatment) yla(0 0.01 "1" 0.02 "2" 0.03 "3" 0.04 "4") ytitle(Percent) name(G2, replace)

If I understand your question correctly, you have a 2 x 2 classification but only 2 out of 4 possibilities are present in your data example. But then I am not sure that I understand your set-up as the percent coming to a walk will be zero for those not coming to a walk.

You may prefer graph bar to graph hbar.

Here is one of those graphs. The other just flips the order of the categories.

[matze pie]

Thank you Mr. Cox for your help. I think i understand that you mean with a 2x2 classification so I correct my data. Gruppe means 1 Persönliche Gruppe 0 Formelle Gruppe. 1 Teilgenommen and 0 nicht teilgenommen means going to a walk or not.

* Example generated by -dataex-. For more info, type help dataex
clear
input byte ID str3 Mitarbeiter byte(Gruppe Teilnahme)
 1 "K01" 1 0
 2 "H02" 1 0
 3 "G03" 1 0
 4 "D04" 1 0
 5 "H05" 1 0
 6 "M06" 1 0
 7 "A07" 1 0
 8 "L08" 1 0
 9 "N09" 1 0
10 "P10" 1 0
11 "D11" 1 0
12 "K12" 1 0
13 "P13" 1 1
14 "B14" 1 0
15 "H15" 1 0
16 "F16" 1 0
17 "E17" 1 0
18 "B18" 1 0
19 "R19" 1 0
20 "K20" 1 0
21 "W21" 1 0
22 "L22" 1 0
23 "M23" 1 0
25 "W25" 0 0
26 "W26" 0 0
27 "B27" 0 0
28 "C28" 0 0
29 "M29" 0 0
30 "S30" 0 0
31 "A31" 0 0
32 "D32" 0 0
33 "B33" 0 0
34 "S34" 0 0
35 "O35" 0 0
36 "M36" 0 0
37 "H37" 0 0
38 "G38" 0 0
39 "B39" 0 1
40 "K40" 0 0
41 "S41" 0 0
42 "H42" 0 0
43 "F43" 0 0
44 "W44" 0 0
45 "D45" 0 0
46 "K46" 0 0
47 "M47" 0 0
48 "B48" 0 0
end

I wanted to use this code but Stata said bargraph_g.new ID, over(Teilnahme) over(Gruppe) title("Teilnahme am Spaziergang nach Gruppe") legend(order(1 "Formelle Gruppe" 2 "Persönliche
Gruppe")): class member function not found and sometimes command legend, title unrecognized

I used this code for the graph bar
* 4. Balkendiagramm
graph bar (count), over(Teilnahme) over(Gruppe) bar(1, color(blue)) bar(2, color(red)) title("Teilnahme am Spaziergang nach Gruppe") legend(label(1 "Formelle Gruppe") label(2 "Persönliche Gruppe"))

[Nick Cox]

I get two colours if I add the asyvars option. The command you give is legal and works. The problems you report must arise with code you don't show us, perhaps because you are splitting a command line.

* Example generated by -dataex-. For more info, type help dataex
clear
input byte ID str3 Mitarbeiter byte(Gruppe Teilnahme)
 1 "K01" 1 0
 2 "H02" 1 0
 3 "G03" 1 0
 4 "D04" 1 0
 5 "H05" 1 0
 6 "M06" 1 0
 7 "A07" 1 0
 8 "L08" 1 0
 9 "N09" 1 0
10 "P10" 1 0
11 "D11" 1 0
12 "K12" 1 0
13 "P13" 1 1
14 "B14" 1 0
15 "H15" 1 0
16 "F16" 1 0
17 "E17" 1 0
18 "B18" 1 0
19 "R19" 1 0
20 "K20" 1 0
21 "W21" 1 0
22 "L22" 1 0
23 "M23" 1 0
25 "W25" 0 0
26 "W26" 0 0
27 "B27" 0 0
28 "C28" 0 0
29 "M29" 0 0
30 "S30" 0 0
31 "A31" 0 0
32 "D32" 0 0
33 "B33" 0 0
34 "S34" 0 0
35 "O35" 0 0
36 "M36" 0 0
37 "H37" 0 0
38 "G38" 0 0
39 "B39" 0 1
40 "K40" 0 0
41 "S41" 0 0
42 "H42" 0 0
43 "F43" 0 0
44 "W44" 0 0
45 "D45" 0 0
46 "K46" 0 0
47 "M47" 0 0
48 "B48" 0 0
end

graph bar (count), over(Teilnahme) over(Gruppe) asyvars bar(1, color(blue)) bar(2, color(red)) title("Teilnahme am Spaziergang nach Gruppe") legend(label(1 "Formelle Gruppe") label(2 "Persönliche Gruppe"))

[matze pie]

Hallo and thank you Mr. Cox . Here is my complete Code. I had to do fisher exact Test and nothing else before.

* 1. Lade die Daten
import excel "Hierist der OrdnerPfad", sheet("Tabelle1") firstrow clear

* 2. Überprüfe die geladenen Variablen
describe

* 3. Fisher-Exakt-Test durchführen
tabulate Gruppe_Persoenlich Spaziergang, exact

* 4. Balkendiagramm
graph bar (count), over(Teilnahme) over(Gruppe) bar(1, color(blue)) bar(2, color(red)) title("Teilnahme am Spaziergang nach Gruppe") legend(label(1 "Formelle Gruppe") label(2 "Persönliche Gruppe"))

[Nick Cox]

I can't run your code as I don't have your spreadsheet. As said, the problems mentioned in

I wanted to use this code but Stata said bargraph_g.new ID, over(Teilnahme) over(Gruppe) title("Teilnahme am Spaziergang nach Gruppe") legend(order(1 "Formelle Gruppe" 2 "Persönliche
Gruppe")): class member function not found and sometimes command legend, title unrecognized

must arise from code you don't show us.

[matze pie]

Hallo Mr. Cox I fix my Problem. I installed Stata on an other Laptop and it works to plot graphs. Now I have a new Problem with titles . Here my complete Code

* Berechnung der Gesamtanzahl pro Gruppe
egen Gesamt = count(Teilnahme), by(Gruppe)

* Berechnung der Anzahl der Teilnehmer pro Gruppe
egen Teilnehmer = total(Teilnahme), by(Gruppe)

* Berechnung des prozentualen Anteils der Teilnehmer
gen Teilnahme_Prozent = (Teilnehmer / Gesamt) * 100 if Teilnahme == 1

graph bar Teilnahme_Prozent, over(Gruppe, label(labsize(large))) bar(1, color(blue))
> blabel(bar, format(%6.1f)) title("Teilnahme am Spaziergang nach Gruppe") ylabel(0(1
> 0)100, labsize(large)) ytitle("Prozentuale Teilnahme (%)", labsize(large)) xlabel(0
> "Formelle Gruppe" 1 "Persönliche Gruppe", labsize(large))
xlabels(0 Formelle not found


I want to rename my y to "Teilnahme in Prozent" and x for 0 "FormelleGruppe" and 1 "Persönliche Gruppe". Thank you vor helping me

[Chen Samulsion]

Options xlabel(), ..., xmtick() are irrelevant for bar charts.

label define Gruppe 0 "Formelle Gruppe" 1 "Persönliche Gruppe"
label values Gruppe Gruppe
graph bar Teilnahme_Prozent, over(Gruppe, label(labsize(large))) bar(1, color(blue)) blabel(bar, format(%6.1f)) title("Teilnahme am Spaziergang nach Gruppe") ylabel(0(10)100, labsize(large)) ytitle("Teilnahme in Prozent(%)")

[Nick Cox]

The suboption labsize() doesn't apply to title options. graph bar doesn't consider the categorical axis to be an x axis.

This may help:

clear
input byte ID str3 Mitarbeiter byte(Gruppe Teilnahme)
 1 "K01" 1 0
 2 "H02" 1 0
 3 "G03" 1 0
 4 "D04" 1 0
 5 "H05" 1 0
 6 "M06" 1 0
 7 "A07" 1 0
 8 "L08" 1 0
 9 "N09" 1 0
10 "P10" 1 0
11 "D11" 1 0
12 "K12" 1 0
13 "P13" 1 1
14 "B14" 1 0
15 "H15" 1 0
16 "F16" 1 0
17 "E17" 1 0
18 "B18" 1 0
19 "R19" 1 0
20 "K20" 1 0
21 "W21" 1 0
22 "L22" 1 0
23 "M23" 1 0
25 "W25" 0 0
26 "W26" 0 0
27 "B27" 0 0
28 "C28" 0 0
29 "M29" 0 0
30 "S30" 0 0
31 "A31" 0 0
32 "D32" 0 0
33 "B33" 0 0
34 "S34" 0 0
35 "O35" 0 0
36 "M36" 0 0
37 "H37" 0 0
38 "G38" 0 0
39 "B39" 0 1
40 "K40" 0 0
41 "S41" 0 0
42 "H42" 0 0
43 "F43" 0 0
44 "W44" 0 0
45 "D45" 0 0
46 "K46" 0 0
47 "M47" 0 0
48 "B48" 0 0
end

egen Gesamt = count(Teilnahme), by(Gruppe)
egen Teilnehmer = total(Teilnahme), by(Gruppe)

gen Teilnahme_Prozent = (Teilnehmer / Gesamt) * 100 if Teilnahme == 1

graph bar Teilnahme_Prozent, over(Gruppe, relabel(1 "Formelle Gruppe" 2 "Persönliche Gruppe")  ///
label(labsize(large))) bar(1, color(blue)) ///
blabel(bar, format(%6.1f)) title("Teilnahme am Spaziergang nach Gruppe") ///
ylabel(0(10)100, labsize(large)) ytitle("Prozentuale Teilnahme (%)", size(large))

[Nick Cox]

Chen Samulsion wrote

Options xlabel(), ..., xmtick() are irrelevant for bar charts.

True for graph bar and graph hbar. Not true for twoway bar etc.

[matze pie]

Thank you for helping me, Mr. Cox and Mr. Samulsion. Now I have the graph I need.

[Meng Zhang]

Dear Professor Cox,

I hope this message finds you well. I am currently working on generating a bar graph in Stata and have encountered an issue with the axis labels. The figure I have attached illustrates the desired outcome: displaying bars for each year but showing x-axis labels only for specific years (e.g., 2007, 2010, 2013, etc.).

In my current approach, I am using the following code:


graph bar (asis) share_other share_same if eic==1, ///
over(year) stack ///
bar(1, color(blue%60)) bar(2, color(green%30)) ///
legend(order(1 "Other Cities" 2 "Within Cities") size(small) region(lcolor(white))) ///
ytitle("Percentage, %", size(small)) ylabel(0(20)100, labsize(small)) ///
graphregion(fcolor(white))

However, this results in all years being displayed on the x-axis (i.e., 2007, 2008, 2009, etc.), leading to a cluttered appearance. I was able to achieve the desired labeling manually through the "Add or customize individual ticks and labels (major)" option in the Graph Editor, as shown in the second attached figure.

I am seeking a way to replicate this selective labeling programmatically within the code. Could you please advise on how to modify the Stata code to display bars for all years but include x-axis labels only for specific years, such as 2007, 2010, 2013, and so on?

Thank you for your time and assistance.

Best regards,
Meng

[Nick Cox]

See https://journals.sagepub.com/doi/pdf...6867X211000032 for a detailed discussion of your problem.

[Meng Zhang]

Thanks for your quick reply—it worked perfectly. I used this code:

label val year
local call
forvalues j = 2(2)20 {
local call `call' `j' " "
}

graph bar (asis) share_other share_same if eic==1, ///
over(year, relabel(`call')) stack ///
bar(1, color(blue%60)) bar(2, color(green%30)) ///
legend(order(1 "Other Cities" 2 "Within Cities") size(small) region(lcolor(white))) ///
ytitle("Percentage, %", size(small)) ylabel(0(20)100, labsize(small)) ///
graphregion(fcolor(white))

The result looks much better. However, I noticed a minor issue: the x-axis now labels 2007, 2009, 2011, etc., but I’d like it to show 2007, 2010, 2013, and so on.

[Nick Cox]

Your years are 2007 to 2019 in 13 bars and you want to blank out labels for bars (2 3) (5 6) (8 9) (11 12) so that you only get labels for bars 1 4 7 10 13. So
the code could be

local call

forvalues j = 2(3)11 {
local call `call' `j' ""
local k = `j' + 1
local call `call' `k' ""
}

So first time around the loop you blank out the label for bar 2 -- and also for bar 3. Then 5 and 6, and so on.