Using labels of variable as a title of line plot with tsline

[Tariq Masood]

In this code I am trying to use labels of variable Sector_ID as titles of individual plots, kindly guide: ************************************************** **

local graphnames
levelsof Sector_ID, local(categories)
foreach c of local categories {
tsline EE if Sector_ID == `c', name(graph`c', replace) ///
title("`c'") ///
ytitle("") xtitle("") ///
local graphnames "`graphnames' graph`c'"
}

[Andrew Musau]

webuse grunfeld, clear
keep if company==1
lab var invest "Gross investment in 1947 dollars"
lab var mvalue  "Market value as of Dec. 31 in 1947 dollars"
lab var kstock   "Stock of plant and equipment in 1947 dollars"

foreach var of varlist invest mvalue kstock{
    tw line `var' year, sort xtitle("") ytitle("") title("`:var lab `var''") saving(gr_`var', replace)
}
gr combine gr_invest.gph gr_mvalue.gph gr_kstock.gph

[Nick Cox]

I guess that Tariq Masood wants to see string values (if Sector_ID is string) or value labels (if Sector_ID is numeric).

Tariq's code should run without error if the identifier is a numeric variable, but you won't see value labels.

So, Tariq, which is it? This is why we ask for data examples. Your question has two quite different solutions depending which type of variable you have.

(In any case why loop at all?

tsline EE, by(Sector_ID)

is a cleaner way to get those graphs.)

[Tariq Masood]

Dear Nick Cox and Andrew Musau I am pasting data and commands here,
* Example generated by -dataex-. For more info, type help dataex
clear
input byte(Time_ID Sector_ID) float EE
1 1 .3770609
2 1 .37964424
3 1 .382227
4 1 .3848091
5 1 .3873904
6 1 .3899709
7 1 .3925503
8 1 .3951286
9 1 .3977056
10 1 .4002813
11 1 .4028554
12 1 .405428
13 1 .4079989
14 1 .41056785
15 1 .4131349
16 1 .4156999
17 1 .4182628
18 1 .4208234
19 1 .4233815
20 1 .4259372
21 1 .4284903
22 1 .4310407
23 1 .4335883
24 1 .436133
25 1 .4386747
26 1 .44121325
27 1 .4437486
28 1 .4462806
29 1 .4488093
30 1 .4513345
31 1 .453856
32 1 .4563739
33 1 .458888
34 1 .4613982
35 1 .4639045
36 1 .4664067
37 1 .4689048
38 1 .4713986
1 2 .0017546547
2 2 .0019141063
3 2 .002081986
4 2 .002258339
5 2 .002443201
6 2 .002636599
7 2 .002838551
8 2 .003049071
9 2 .003268168
10 2 .0034958455
11 2 .003732107
12 2 .0039769537
13 2 .0042303856
14 2 .0044924053
15 2 .004763014
16 2 .005042217
17 2 .00533002
18 2 .005626434
19 2 .005931473
20 2 .006245153
21 2 .006567495
22 2 .006898525
23 2 .007238273
24 2 .007586773
25 2 .007944067
26 2 .008310194
27 2 .008685206
28 2 .009069155
29 2 .0094621
30 2 .0098641
31 2 .010275223
32 2 .010695542
33 2 .011125128
34 2 .01156406
35 2 .01201242
36 2 .012470296
37 2 .012937773
38 2 .013414945
1 3 .014721413
2 3 .015329368
3 3 .015948879
4 3 .016579906
5 3 .017222416
6 3 .01787639
7 3 .018541818
8 3 .0192187
9 3 .01990703
10 3 .02060683
11 3 .021318117
12 3 .02204092
13 3 .022775274
14 3 .02352122
15 3 .0242788
16 3 .025048066
17 3 .025829077
18 3 .02662189
19 3 .027426576
20 3 .0282432
21 3 .02907183
22 3 .02991255
23 3 .030765435
24 3 .031630557
end
label values Sector_ID Sector_ID
label def Sector_ID 1 "Agriculture", modify
label def Sector_ID 2 "Mining", modify
label def Sector_ID 3 "Food & Bev", modify
label var Time_ID "Time_ID"
label var Sector_ID "Sector_ID"

xtset Sector_ID Time_ID

// Create a local macro to store the graph names
local graphnames ""

// Loop over the unique values of 'cat'
levelsof Sector_ID, local(categories)
foreach c of local categories {
// Create a time series line graph for each category
tsline EE if Sector_ID == `c', name(graph`c', replace) ///
title("`c'") ///
ytitle("") xtitle("") ///
// Add the graph name to the local macro
local graphnames "`graphnames' graph`c'"
}

// Combine the graphs into a single graph
graph combine `graphnames', rows(1) cols(3) ///
title("Dynamics of Total Factor Energy Efficiency")


By running this, I am getting a combined graph with titles "1", "2" and "3" while I want them as "Agriculture", "Mining" and "Food & Bev" which are labels of variable Sector_id.

by
tsline EE, by(Sector_ID) the problem is all Y axes of all three categories have the same axis values, which hide the sectoral variability in the graph.

[Nick Cox]

Thanks for the example and code. You should want to look up the value label but leave the numbers as graph names (as spaces and & can't be part of Stata names)

Not so to the last problem, as an option exists for precisely this situation. See the last command below, which gets you to a good place directly.

clear
input byte(Time_ID Sector_ID) float EE
1 1 .3770609
2 1 .37964424
3 1 .382227
4 1 .3848091
5 1 .3873904
6 1 .3899709
7 1 .3925503
8 1 .3951286
9 1 .3977056
10 1 .4002813
11 1 .4028554
12 1 .405428
13 1 .4079989
14 1 .41056785
15 1 .4131349
16 1 .4156999
17 1 .4182628
18 1 .4208234
19 1 .4233815
20 1 .4259372
21 1 .4284903
22 1 .4310407
23 1 .4335883
24 1 .436133
25 1 .4386747
26 1 .44121325
27 1 .4437486
28 1 .4462806
29 1 .4488093
30 1 .4513345
31 1 .453856
32 1 .4563739
33 1 .458888
34 1 .4613982
35 1 .4639045
36 1 .4664067
37 1 .4689048
38 1 .4713986
1 2 .0017546547
2 2 .0019141063
3 2 .002081986
4 2 .002258339
5 2 .002443201
6 2 .002636599
7 2 .002838551
8 2 .003049071
9 2 .003268168
10 2 .0034958455
11 2 .003732107
12 2 .0039769537
13 2 .0042303856
14 2 .0044924053
15 2 .004763014
16 2 .005042217
17 2 .00533002
18 2 .005626434
19 2 .005931473
20 2 .006245153
21 2 .006567495
22 2 .006898525
23 2 .007238273
24 2 .007586773
25 2 .007944067
26 2 .008310194
27 2 .008685206
28 2 .009069155
29 2 .0094621
30 2 .0098641
31 2 .010275223
32 2 .010695542
33 2 .011125128
34 2 .01156406
35 2 .01201242
36 2 .012470296
37 2 .012937773
38 2 .013414945
1 3 .014721413
2 3 .015329368
3 3 .015948879
4 3 .016579906
5 3 .017222416
6 3 .01787639
7 3 .018541818
8 3 .0192187
9 3 .01990703
10 3 .02060683
11 3 .021318117
12 3 .02204092
13 3 .022775274
14 3 .02352122
15 3 .0242788
16 3 .025048066
17 3 .025829077
18 3 .02662189
19 3 .027426576
20 3 .0282432
21 3 .02907183
22 3 .02991255
23 3 .030765435
24 3 .031630557
end
label values Sector_ID Sector_ID
label def Sector_ID 1 "Agriculture", modify
label def Sector_ID 2 "Mining", modify
label def Sector_ID 3 "Food & Bev", modify
label var Time_ID "Time_ID"
label var Sector_ID "Sector_ID"

xtset Sector_ID Time_ID

// Create a local macro to store the graph names
local graphnames ""

// Loop over the unique values of 'cat'
levelsof Sector_ID, local(categories)
foreach c of local categories {
local ctext : label (Sector_ID) `c'
// Create a time series line graph for each category
tsline EE if Sector_ID == `c', name(graph`c', replace) ///
title("`ctext'") ///
ytitle("") xtitle("") ///
// Add the graph name to the local macro
local graphnames "`graphnames' graph`c'"
}

// Combine the graphs into a single graph
graph combine `graphnames', rows(1) cols(3) ///
title("Dynamics of Total Factor Energy Efficiency") name(graphALL1, replace)


tsline EE, by(Sector_ID, yrescale row(1) note("")) name(graphALL2, replace)

[Tariq Masood]

Thanks Nick. Now it perfectly works.