Hello everyone,
Ιn the scatter plot below, I calculate the time effects of a categorical variable in terms of counts in time (1-15) for local councils on some economic aspects for a panel in different countries. Local councils are renewing their period every 4–5 years, depending on the case, with some rare cases having a longer duration of up to 15 years, in irregular countries (i.e.same leader regime, lower degree of democracy, or iconic democracy). That means there are more observations in years 4-5 when the renewal is happening than later years, from 6 up to the year 15, and there are just a few of them. However, taking a graph of summary statistics for year 15 or so, in some variables, the mean statistic is higher, since for instance, there are only 16 observations in year 15 with respect to 500+ observations in years 1–5, which is misleading. According to the model, there should be an increase(decrease) in the mean in the first year, dropping afterward, subject to the case, the variable economic meaning and sign effect. In other words, the early years are far more effective. That means there should be a short of pattern there, if a close look in the scatter plot and the tabulation below.
I do not mean to winsorize that, but somehow to treat it, creating a graph and a summary table, showing the pattern as well as a correlation table. It is hard to provide data as I have data from more than 200 local councils, and they reach up to 4000 observations. Hard to find where those few observations are, hard to get the right combination of data in order to provide an example. But if you suggest a way, I could provide a sample example. How do I take care of that?
My scatter code and plot are those below.
Thanks in advance
Giorgio
I also run a summary statistics for the variable count, for example for one of the variables in the scatter.
Ιn the scatter plot below, I calculate the time effects of a categorical variable in terms of counts in time (1-15) for local councils on some economic aspects for a panel in different countries. Local councils are renewing their period every 4–5 years, depending on the case, with some rare cases having a longer duration of up to 15 years, in irregular countries (i.e.same leader regime, lower degree of democracy, or iconic democracy). That means there are more observations in years 4-5 when the renewal is happening than later years, from 6 up to the year 15, and there are just a few of them. However, taking a graph of summary statistics for year 15 or so, in some variables, the mean statistic is higher, since for instance, there are only 16 observations in year 15 with respect to 500+ observations in years 1–5, which is misleading. According to the model, there should be an increase(decrease) in the mean in the first year, dropping afterward, subject to the case, the variable economic meaning and sign effect. In other words, the early years are far more effective. That means there should be a short of pattern there, if a close look in the scatter plot and the tabulation below.
I do not mean to winsorize that, but somehow to treat it, creating a graph and a summary table, showing the pattern as well as a correlation table. It is hard to provide data as I have data from more than 200 local councils, and they reach up to 4000 observations. Hard to find where those few observations are, hard to get the right combination of data in order to provide an example. But if you suggest a way, I could provide a sample example. How do I take care of that?
My scatter code and plot are those below.
Thanks in advance
Giorgio
Code:
scatter gdp_capita_r count, graphr(c(white)) m(Oh) yline(0, lc(navy) lp(dash)) || lfit gdp_capita_r count ///
|| scatter gdp_capita_r count, mc(navy%10) mlc(none) ml() mlabc(navy) mlabpos(3) ///
xtitle("Count in time") yt("") title(" GDP per capita ", size(medsmall)) ///
legend(off) yl(, format(%03.1f))
Code:
-> count = 1
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 173 30128.29 21966.47 581.2183 79299.84
--------------------------------------------------------------------------------
-> count = 2
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 405 24231.69 14149.85 345.4215 61174.54
--------------------------------------------------------------------------------
-> count = 3
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 401 27788.38 15816.52 517.717 81653.34
--------------------------------------------------------------------------------
-> count = 4
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 465 25782.95 19592.54 364.7466 105264.8
--------------------------------------------------------------------------------
-> count = 5
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 203 23246.8 26436.19 330.2053 111968.4
--------------------------------------------------------------------------------
-> count = 6
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 21 1492.847 1629.38 328.0719 4139.031
--------------------------------------------------------------------------------
-> count = 7
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 8 13650.78 2329.838 10347.18 16791.89
--------------------------------------------------------------------------------
-> count = 9
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 10 394.2766 13.27976 373.5514 415.7148
--------------------------------------------------------------------------------
-> count = 11
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 12 599.7394 71.06272 491.9627 713.6509
--------------------------------------------------------------------------------
-> count = 15
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 14 393.3982 59.90866 302.0926 469.4377
--------------------------------------------------------------------------------
-> count = .
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
gdp_capita_r | 560 20751.71 23475.95 322.3344 111043.5
Comment