Propensity score and Mahalanobis matching
by Paolo Chirico
Dear all,
a new package is available on the server: stat_match.
It allows gretl users to apply two methods of statistical matching,
radius and one-to-one methods, using both the propensity score and the
mahalanobis distance.
I will be grateful to all those who send me suggestions to improve it or
point out any bugs.
Cheers,
Paolo
--
Paolo Chirico
RU e Prof.Agg. di Statistica Economica
Università del Piemonte Orientale
Dip. di Giurisprudenza e Scienze Politiche,
Economiche e Sociali (DIGSPES)
Alessandria, Italia
1 year, 5 months
$workdir inheritance
by Marcin Błażejowski
Hi,
I'd like to as if $workdir inheritance is conditional (on something?) or
unconditional. I'm asking because on my machine - which run under Debian
testing/unstable - $workdir is always set to current dir. But on two
Ubuntu-s I have (in Oracle cloud) $workdir is _always_ set to $HOME.
Of course the workaround is trivial via 'getenv("PWD")', but I'd like to
know.
Marcin
--
Marcin Błażejowski
1 year, 5 months
Re: 100-based indices with panel data (fwd)
by Riccardo (Jack) Lucchetti
---------- Forwarded message ----------
Date: Sat, 1 Jul 2023 13:08:04 +0000
From: Allin Cottrell <cottrell(a)wfu.edu>
To: Josué Martínez-Castillo <jota3mc(a)gmail.com>
Subject: Re: 100-based indices with panel data
On Fri, 30 Jun 2023, Josué Martínez-Castillo wrote:
> Hi,
> I'm a newbie in gretl, very excited to learn how to use the program for
> learning econometrics on my own. However, right now I'm curious on how to
> estimate 100-based indices when dealing with panel data. For example, what
> if I want to estimate a 100-based index for each unit using as base year
> the first year available of, say, real GDP.
>
> I was looking for the answer in the manual of the 2023 version of gretl. No
> success. I was hoping maybe someone can help me with guidance.
Good question. As things stand there isn't a built-in way to construct such
indices for panel data using the graphical interface. But assuming you want the
indices to work in the time dimension for each panel unit, it's actually not
hard to do via scripting. Here's an example:
set verbose off
# load a supplied panel dataset
open grunfeld
# add a series to hold indices for the "invest" series
series invest_idx
# determine the number of panel units
N = max($unit)
# loop across the units
loop i=1..N
# restrict the sample range to unit i
smpl i i --unit
# calculate a portion of the full series
invest_idx = 100*invest/invest[$i:1]
endloop
# restore the full sample and inspect the result
smpl full
print invest_idx --byobs
In the example the base for the indices is the first observation for each unit
(unit $i, observation 1) but you could choose any period as the base.
We should probably find a way to offer this as a built-in thing.
Allin Cottrell
1 year, 5 months
Re: 100-based indices with panel data (fwd)
by Riccardo (Jack) Lucchetti
---------- Forwarded message ----------
Date: Sat, 1 Jul 2023 19:39:27 +0000
From: Allin Cottrell <cottrell(a)wfu.edu>
To: "Riccardo (Jack) Lucchetti" <r.lucchetti(a)univpm.it>
Cc: Josué Martínez-Castillo <jota3mc(a)gmail.com>,
Subject: Re: 100-based indices with panel data
On Sat, 1 Jul 2023, Riccardo (Jack) Lucchetti wrote:
> On Sat, 1 Jul 2023, Allin Cottrell wrote:
>
>> On Fri, 30 Jun 2023, Josué Martínez-Castillo wrote:
>>
>>> I'm a newbie in gretl, very excited to learn how to use the program for
>>> learning econometrics on my own. However, right now I'm curious on how to
>>> estimate 100-based indices when dealing with panel data. For example, what
>>> if I want to estimate a 100-based index for each unit using as base year
>>> the first year available of, say, real GDP.
>>>
>>> I was looking for the answer in the manual of the 2023 version of gretl. No
>>> success. I was hoping maybe someone can help me with guidance.
>>
>> Good question. As things stand there isn't a built-in way to construct such
>> indices for panel data using the graphical interface. But assuming you want
>> the indices to work in the time dimension for each panel unit, it's actually
>> not hard to do via scripting. Here's an example:
> [...]
>
> Here's another approach, which avoids the loop. The syntax is a bit too
> terse, perhaps, but IMO instructive.
>
> <hansl>
> open abdata.gdt
>
> base = cum(ok(EMP)) == 1 ? EMP : NA
> EMP_b100 = EMP/pexpand({base}) * 100
> print EMP EMP_b100 --byobs
> </hansl>
Yes, quite instructive! In case anyone's interested let's unpack Jack's
formulation.
First consider:
base = cum(ok(EMP)) == 1 ? EMP : NA
We're looking at what gretl calls series here.
The inner expression, "ok(EMP)" creates a series with value 1 for valid values
of its series argument and 0 for NAs (missing values).
This addresses a problem with the first variant I posted, where I just took the
base of the indices to be the first observation for each unit. That's OK with
the grunfeld data that I referenced because it has no missing values. But if
the first observation for a unit were NA, the whole index series for that unit
would be NA via my method (since NAs propagate in arithmetical calculation).
Not accidentally, Jack chose the supplied abdata dataset (Arellano and Bond),
which contains missing values, to illustrate his calculation, and I'll work
with it here.
Now cum() is gretl's cumulation function, and it works "properly" for panel
data: it cumulates in the time dimension, starting over for each unit. So
"cum(ok(EMP))" gives a series holding the count of valid values "to date" for
each unit. OK, so far?
Then "cum(ok(EMP)) == 1 ? EMP : NA" is an instance of the very handy ternary
operator. It has the form:
result = condition ? one_thing : other_thing
which can be spelled out a bit as
if (condition is true) result is one_thing, otherwise result is other_thing
So, "cum(ok(EMP)) == 1 ? EMP : NA" gives a series holding the value of EMP for
each first-valid-observation per panel unit, and NA for all other observations.
To see what's happening up to this point one could open the abdata dataset in
gretl and execute these commands (in a script or via the console):
series eok = ok(EMP)
series cumeok = cum(eok)
series base = cum(ok(EMP)) == 1 ? EMP : NA
print EMP eok cumeok base --byobs
Next comes the line:
EMP_b100 = EMP/pexpand({base}) * 100
On the left-hand side is the final indices series. On the right-hand side we're
using the original EMP (employment), multiplying by 100 (as per convention),
and dividing by "pexpand({base})". What the heck is this last thing?
Well, notice the curly brackets around "base". These turn a series into a
vector (special case of a matrix) and the thing you need to know here is that
in gretl by default this conversion skips any missing values. [Note: you can
prevent this via the command "set skip_missing off".] So in a panel with N
units {base} will be an N-vector holding just the first valid observation of
EMP for each unit.
Then the pexpand ("panel-expand") function turns this N-vector into a series by
repeating each of the N values T times, for each unit. Which is (probably) just
what we want to divide EMP by, to create the per-unit indices. In a panel
dataset with no missing values it's exactly equivalent to the more pedestrian
formulations I posted earlier.
Now for a couple of missing-data complications we'd want to deal with in a
built-in version of this functionality.
1) What if some units have NO valid values for the variable we're working with?
Then {base} will not be an N-vector and Jack's method will not work unmodified.
2) What if the date of the first valid observation differs across units, but we
want a set of indices that start in the same period? Again, some fancier
footwork would be needed. We'd need to look for the first period with a common
non-missing observation across all units that had more than one non-missing
observation.
Allin Cottrell
1 year, 5 months
Re: 100-based indices with panel data (fwd)
by Riccardo (Jack) Lucchetti
---------- Forwarded message ----------
Date: Sat, 1 Jul 2023 17:36:11 +0000
From: "Riccardo (Jack) Lucchetti" <p002264(a)staff.univpm.it>
Reply-To: "Riccardo (Jack) Lucchetti" <r.lucchetti(a)univpm.it>
To: Allin Cottrell <cottrell(a)wfu.edu>
Cc: Josué Martínez-Castillo <jota3mc(a)gmail.com>,
Subject: Re: 100-based indices with panel data
On Sat, 1 Jul 2023, Allin Cottrell wrote:
> On Fri, 30 Jun 2023, Josué Martínez-Castillo wrote:
>
>> Hi,
>> I'm a newbie in gretl, very excited to learn how to use the program for
>> learning econometrics on my own. However, right now I'm curious on how to
>> estimate 100-based indices when dealing with panel data. For example, what
>> if I want to estimate a 100-based index for each unit using as base year
>> the first year available of, say, real GDP.
>>
>> I was looking for the answer in the manual of the 2023 version of gretl. No
>> success. I was hoping maybe someone can help me with guidance.
>
> Good question. As things stand there isn't a built-in way to construct such
> indices for panel data using the graphical interface. But assuming you want
> the indices to work in the time dimension for each panel unit, it's actually
> not hard to do via scripting. Here's an example:
[...]
Here's another approach, which avoids the loop. The syntax is a bit too terse,
perhaps, but IMO instructive.
<hansl>
open abdata.gdt
base = cum(ok(EMP)) == 1 ? EMP : NA
EMP_b100 = EMP/pexpand({base}) * 100
print EMP EMP_b100 --byobs
</hansl>
> We should probably find a way to offer this as a built-in thing.
Agreed.
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
1 year, 5 months
Re: 100-based indices with panel data (fwd)
by Riccardo (Jack) Lucchetti
---------- Forwarded message ----------
Date: Sat, 1 Jul 2023 14:41:57 +0000
From: Allin Cottrell <cottrell(a)wfu.edu>
To: Josué Martínez-Castillo <jota3mc(a)gmail.com>
Subject: Re: 100-based indices with panel data
On Sat, 1 Jul 2023, Allin Cottrell wrote:
> On Fri, 30 Jun 2023, Josué Martínez-Castillo wrote:
>
>> Hi,
>> I'm a newbie in gretl, very excited to learn how to use the program for
>> learning econometrics on my own. However, right now I'm curious on how to
>> estimate 100-based indices when dealing with panel data. For example, what
>> if I want to estimate a 100-based index for each unit using as base year
>> the first year available of, say, real GDP.
>>
>> I was looking for the answer in the manual of the 2023 version of gretl. No
>> success. I was hoping maybe someone can help me with guidance.
>
> Good question. As things stand there isn't a built-in way to construct such
> indices for panel data using the graphical interface. But assuming you want
> the indices to work in the time dimension for each panel unit, it's actually
> not hard to do via scripting. Here's an example [...]
It might be worth noting that you can get the same result with a good deal
fewer commands by exploiting gretl's matrix methods:
set verbose off
open grunfeld
# turn "invest" into a T x N matrix
matrix m = mshape({invest}, $pd, max($unit))
# scale @m and turn it back into a series
series invest_idx = mshape(100 * m ./ m[1,], $nobs, 1)
print invest_idx --byobs
Allin Cottrell
1 year, 5 months
100-based indices with panel data (fwd)
by Riccardo (Jack) Lucchetti
Hi all.
I'm forwarding to this list a series of messages that Allin and I
exchanged with a Josué, a new user (welcome, by the way), offlist.
---------- Forwarded message ----------
Date: Sat, 1 Jul 2023 03:54:14 +0000
From: Josué Martínez-Castillo <jota3mc(a)gmail.com>
Subject: 100-based indices with panel data
Hi,
I'm a newbie in gretl, very excited to learn how to use the program for
learning econometrics on my own. However, right now I'm curious on how to
estimate 100-based indices when dealing with panel data. For example, what
if I want to estimate a 100-based index for each unit using as base year
the first year available of, say, real GDP.
I was looking for the answer in the manual of the 2023 version of gretl. No
success. I was hoping maybe someone can help me with guidance.
Thanks.
jota3mc
1 year, 5 months