Collect layout error messages (e.g., collect dims not returning colnames among other things)

[Emilya Hennessy]

Hello, I'm new to using the collect functionality in Stata and have been working my way through the user manual, but cannot reproduce most of the initial output with my own data. In my syntax below, sex is a binary variable and subject_age is continuous. I can get some code to work but not others even though I (think) I am following the examples exactly. I'm trying to eventually build out larger tables but wanted to get the basics down before trying that.

Thank you!
Emily



. collect clear

. sort sex

. collect: by sex:summarize subject_age

-------------------------------------------------------------------------------------------------------------------------------------------------------------------
-> sex = Female

Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
subject_age | 42 17.42857 1.532459 13 19

-------------------------------------------------------------------------------------------------------------------------------------------------------------------
-> sex = Male

Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
subject_age | 66 16.95455 1.533541 13 19

. collect layout (subject_age) (sex) (result[mean])
(dimension subject_age not found)

Collection: default
Rows: subject_age
Columns: sex
Tables: result[mean]
Table 1: 1 x 2

-----------------
Female Male
-----------------
17.42857 16.95455
-----------------

. collect dims

Collection dimensions
Collection: default
-----------------------------------------
Dimension No. levels
-----------------------------------------
Layout, style, header, label
cmdset 2
program_class 2
result 8
result_type 1
sex 2

Style only
border_block 4
cell_type 4
-----------------------------------------




collect clear

. sort sex

. collect: by sex:summarize subject_age

-------------------------------------------------------------------------------------------------------------------------------------------------------------------
-> sex = Female

Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
subject_age | 42 17.42857 1.532459 13 19

-------------------------------------------------------------------------------------------------------------------------------------------------------------------
-> sex = Male

Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
subject_age | 66 16.95455 1.533541 13 19

. collect layout (subject_age#sex) (result)
(dimension subject_age not found)

Collection: default
Rows: subject_age#sex
Columns: result

Your layout specification does not uniquely match any items. One or both of the dimensions cmdset and sex might help uniquely match items.


. collect label list cmdset, all

Collection: default
Dimension: cmdset
Label: Command results index
Level labels:
1
2


. collect layout (subject_age#sex) (cmdset)

Collection: default
Rows: subject_age#sex
Columns: cmdset

Your layout specification does not uniquely match any items. One or both of the dimensions result and sex might help uniquely match items.

[Andrew Musau]

There is no dimension named "subject_age". Just leave that empty.

collect clear
sort sex
collect: by sex:summarize subject_age
collect layout () (sex) (result[mean])

[Emilya Hennessy]

Thank you! I missed in the directions that the dimensions are created using the "by" command - so any variables for the interactions described in the instructions (subject_age#sex) first have to have that run. But it won't make sense to "by" a continuous variable for this purpose. I have some work to do!

[Andrew Musau]

You have two variables in your example: age and sex. Do you have a specific table in mind? If you illustrate how you want the table to look, it may be possible to write code that replicates it.