StableFill

Matrix Shapes

Input files must contain rows of numbers or characters preceded by <label>. Rows may be tab-delimited or whitespace-delimited. Tabs are still preferred when text cells contain spaces, but whitespace-delimited rows are convenient for pasted regression output. The numbers can be arbitrarily long, can be negative, and can also be in scientific notation.

For economics regression output, a numeric estimate followed immediately by a parenthesized numeric standard error is treated as one cell:

0.125*** (0.031) -0.456 (0.100)

is parsed as two entries, 0.125*** (0.031) and -0.456 (0.100). A row that starts with parenthesized standard errors, such as (0.031) (0.100), is parsed as separate entries because those standard errors usually occupy their own row.

<tab:Test>
1	2	3
2	3	1
3	1	2

The rows do not need to be of equal length.

<tab:FunnyMat>
1	2	3	23	2
2	3
3	1	2	2
1

Completely blank (no tab) lines are ignored. If a “cell” is merely “.”, “[space]”, or “NA” then it is treated as missing. That is, in the program:

<tab:Test>
1	2	3
2	.	1	3
3	    1	2

is equivalent to:

<tab:Test>
1	2	3
2	1	3
3	1	2

This feature is useful as several languages output missing values as NA, blank, or “.”. Last, StableFill understands scientific notation of the form: [numbers].[numbers]e(+/-)[numbers]

<tab:TestScientific>
23.2389e+23
-2.23e-2
-0.922e+3

Named Values

Single statistics can be written as <Val:...> or <Value:...> blocks and referenced with inline placeholders:

<Val:sample_size>
5708
The sample includes  observations.