Dear All, there seems to be a problem with exporting quarterly formatted data to Excel from Stata 13.1 (Windows)
Take data from here:
http://www.stata-press.com/data/r7/friedman2.dta
Last variable was added with:
Export to Excel via a dialog with default settings.
Obtain the following:

The constant 21916 is rounded by Excel, but saved with some variation in the fractional part:

Given that the numbers are increasing in column B it seems that the values are somehow distorted original quarterly values. Perhaps a transformation was applied similar to what is described here: http://blog.stata.com/2011/01/05/usi...ther-software/ where it isn't applicable.
Anyhow, it is difficult to work with such an export format, and formatting the data as a number (hence column F corresponding to time_val) makes more sense.
For curious, 21916 is the number of days from Jan 1, 1900 (Excel base date) toJan 1, 1960 (Stata base date).
There was already a problem fixed in 13.0-->13.1 regarding the %d variables:

but it is not clear which formats were covered under the %d designation (just one format or all date formats collectively).
The Stata version I am using is as of: 06 Aug 2014
Is this already a known issue? Is this fixed in the most recent update (I know there is more available)?
Thank you, Sergiy Radyakin
Take data from here:
http://www.stata-press.com/data/r7/friedman2.dta
Last variable was added with:
Code:
generate time_val=time
Obtain the following:
The constant 21916 is rounded by Excel, but saved with some variation in the fractional part:
Given that the numbers are increasing in column B it seems that the values are somehow distorted original quarterly values. Perhaps a transformation was applied similar to what is described here: http://blog.stata.com/2011/01/05/usi...ther-software/ where it isn't applicable.
Anyhow, it is difficult to work with such an export format, and formatting the data as a number (hence column F corresponding to time_val) makes more sense.
For curious, 21916 is the number of days from Jan 1, 1900 (Excel base date) toJan 1, 1960 (Stata base date).
There was already a problem fixed in 13.0-->13.1 regarding the %d variables:
but it is not clear which formats were covered under the %d designation (just one format or all date formats collectively).
The Stata version I am using is as of: 06 Aug 2014
Is this already a known issue? Is this fixed in the most recent update (I know there is more available)?
Thank you, Sergiy Radyakin
Comment