-xttrans- message: "no observations"

[Michael Duarte Goncalves]

Hi everyone,

I have some problems with -xttrans- command and need some help please.
To provide some background, here is a dataset:

* Example generated by -dataex-. For more info, type help dataex
clear
input long(id sp_zipcode) double(date_contract_start date_contract_end) float kW_power_p1 long(product_classification_encod tariff_ekon_id_encod idcontrato) float solar_dummy
74163 7010 20462 21608 3.45 3 2 93357 0
31196 7600 20462 21608 3.45 3 2 93367 0
27112 7400 20465 21608 5.75 3 2 93375 0
17066 7310 20521 20971 5.75 3 2 93376 0
36635 7518 20503 20517 5.75 3 2 93380 0
44639 7570 20467 21608 5.75 3 2 93384 0
35215 7590 20458 21608 5.75 3 2 93387 0
33886 7530 20458 21608 5.75 3 2 93389 0
17068 7340 20461 21608 4.6 3 2 93395 0
35869 7670 20466 21608 9.2 1 2 93398 0
30888 7340 20461 21274 5.75 1 2 93407 0
22556 7620 20458 21307 4.6 1 2 93421 0
27386 7011 20459 21608 3.45 3 2 93425 0
15695 7141 20465 21608 6.9 3 2 93426 0
end
format %td date_contract_start
format %td date_contract_end
abel values product_classification_encod product_classification_encod label def product_classification_encod 1 "Clasico", modify
label def product_classification_encod 3 "Sin Sorpresas", modify
label values tariff_ekon_id_encod tariff_ekon_id_encod
label def tariff_ekon_id_encod 2 "20DHA", modify

Basically, I declare -xtset- as a first option:

 xtset idcontrato date_contract_start

I am interested in knowing the transition probabilities between different states for households when they change their contracts. What is the probability of observing this or that switch in contracted powers ("kW_power_p1"), in tariffs ("tariff_ekon_id"), etc?
I ran for example the code below and obtained the following message in red:

· xttrans kW_power_p1 if !missing(powers_tariff2_less_15000w )
no observations

· xttrans product_classification_encod if !missing(powers_tariff2_less_15000w )
no observations

· xttrans tariff_ekon_id_encod if !missing(powers_tariff2_less_15000w )
no observations

Does anyone know why I obtained this message?
Thank you in advance.

Best,
Michael

[Michael Duarte Goncalves]

I tried to do this code:

xttrans tariff_ekon_id_encod
i() required
r(198);

But gives me this error. I don't know how to proceed.
Best,
Michael

[Carlo Lazzaro]

Michael:
you should -xtset- your dataset first as in the following non-sensical toy-example:

. xtset id date_contract_start

Panel variable: id (weakly balanced)
 Time variable: date_contract_start, 05jan2016 to 08mar2016
         Delta: 1 day

. xttrans tariff_ekon_id_encod
no observations

[Michael Duarte Goncalves]

Hi,

Thank you so much for your feedback. The dataex on #1 does not contain some of my repeated ids. Sorry about that.
I obtain the following error in my "true" dataset:

. xtset id date_contract_start
repeated time values within panel
r(451);

Here is an updated -dataex- which seems more realistic:

* Example generated by -dataex-. For more info, type help dataex
clear
input long(id idcontrato sp_zipcode) double(date_contract_start date_contract_end) long(power_p1 tariff_ekon_id_encod product_classification_encod)
1001    1001  9200 18887 21700 4600 1 1
1001  451697  9200 21701 22431 4600 2 1
1001 1236132  9200 22432 22645 4600 4 1
1001 1730454  9200 22646 22676 4600 4 1
1001 2082075  9200 22677 22735 4600 4 1
1001 2172904  9200 22736 23010 4600 4 1
1001 2872183  9200 23011 23069 4600 4 1
1001 3107888  9200 23070     . 4600 4 4
1005    1005 48600 18800 21639 3300 1 1
1005  420392 48600 21640 21651 3300 1 1
end
format %td date_contract_start
format %td date_contract_end
label values tariff_ekon_id_encod tariff_ekon_id_encod
label def tariff_ekon_id_encod 1 "20A", modify
label def tariff_ekon_id_encod 2 "20DHA", modify
label def tariff_ekon_id_encod 4 "20TD", modify
label values product_classification_encod product_classification_encod
label def product_classification_encod 1 "Clasico", modify
label def product_classification_encod 4 "Tarifa Justa", modify

Thank you again.
Best regards,
Michael

[Michael Duarte Goncalves]

date_contract_start date_contract_end are not given on a daily basis.
Rather, they are rather scattered, and I don't know how to deal with this.

• Is this why it keeps giving me "no observation"?

Because looking at my dataex in the post #4, we can clearly see that we are observing a change in tariffs (from 1 to 2) for id 1001 between May 31, 2019 and June 1, 2019 (contract end: 21700. contract start: 21701) for example.

Best,
Michael

[Carlo Lazzaro]

MIchael:
if you gìhave repeated time values within panel, you can -xtset- your dataset with -panelid- only.
Unfortunately, this fix does not allow you to use-xttrans-.

[Michael Duarte Goncalves]

Hi Lazzaro,

Thank you for your helpful feedback!
That's a pity.

I wish you a wonderful day.
Best,

Michael