On Fri, 24 Jan 2020, klaus.hasenbach(a)web.de wrote:
Gretl with dbnomics works very well for me and I`m very happy with both.
But sometimes it is not possible to retrieve time series which are in
the databank of dbnomics and can be retrieved on their web site correctly.
For instance a regular script "data
ECB/BSI/M.U2.N.R.LRE.X.1.A1.3000.Z01.E" does not work for this time
series in gretl. When I go to the search menu of dbnomics in gretl
instead and try to chart this time series it says "nicht endliche
Werte" what more or less means "not finite values" in english.
Thanks for the report. This is a point we need to raise with the
dbnomics people. The ECB series you mention holds 230 monthly
observations from 1999-02 to 2019-12. But that period includes 251
months: 21 observations are missing. Now there are two ways of
A. Give 251 dates and put "NA" (or similar) for the 21 missing
B. Give only 230 dates (skipping some of them altogther) and leave
it up to the reader to figure where missing values should be
I believe that dbnomics has used method A up till recently, and
that's what gretl is expecting. But these ECB data use method B, so
it appears to us that this is not a valid time series -- the dates
are not consecutive. (Side note: if you download CSV from the
dbnomics site the dates are consecutive, with NAs inserted, but if
you download the JSON, as gretl does, the missing values are just
In principle gretl could handle case B, but it's more complicated
and we need to determine if this can be expected, going forward, or
it's an anomaly that the dbnomics people are willing to fix.