Line plot

[Collin James]

Hello,

I would like to make a plot similar to the picture below. My data is below as well. I want to graph placementdays on the X axis and the percentage on the Y axis. Thanks--CJ

* Example generated by -dataex-. For more info, type help dataex
clear
input float placementdays1
  1
 55
  2
  3
  2
  1
  7
 23
 56
 40
 90
 22
 13
  1
  3
 43
  .
 21
183
843
  4
 14
 10
 22
 26
 78
  8
135
  9
 77
105
  .
369
112
147
 21
  5
574
  3
154
  3
344
  2
  3
623
  4
 62
  3
  3
  2
 14
 26
813
682
  0
  3
138
 55
 30
  5
 63
 17
  3
 10
 57
 32
 50
 78
  1
 55
168
561
 50
 21
 19
  .
  .
 55
337
  5
  6
 21
499
  3
153
170
  6
261
261
 28
  3
277
  2
  1
 81
 12
  2
  1
  2
 85
end

[Collin James]

I guess the issue is that I want to graph the percent that have placementdays==1,2,3,.......500,.....etc but not sure hwo to make this variable?

[Hemanshu Kumar]

Something along these lines?

contract placementdays1, cpercent(cumulperc)
gen float still_in_care = 100 - cumulperc
twoway line still_in_care placementdays1, xtitle("Placement days") ytitle("% still in care")

NB. the scheme used for the graph below is lean1, available by typing

net install gr0002_3.pkg

[Nick Cox]

I am not sure that I understand the graph, which doesn't appear to show monotonic decline.

distplot from the Stata Journal is one command for complementary or reverse cumulative distributions (survivor or survival functions).

. search distplot, historical

Search of official help files, FAQs, Examples, and Stata Journals

FAQ     . . . . . . . . . . . . . . . . . . . . . . . . . . .  NJC Stata plots
        . . . . . . . . . . . . . . . . . .  UCLA Academic Technology Services
        7/08    https://stats.idre.ucla.edu/stata/faq/graph/njc-stata-plots/

SJ-19-1 gr41_5  . . . . . . . . . . . . . . . . . Software update for distplot
        (help distplot if installed)  . . . . . . . . . . . . . . .  N. J. Cox
        Q1/19   SJ 19(1):260
        changes include better handling of the by() option calls;
        simpler default y-axis titles; more detailed discussion of
        exactly what is plotted; and more information on ridits

SJ-10-1 gr41_4  . . . . . . . . . . . . . . . . . Software update for distplot
        (help distplot if installed)  . . . . . . . . . . . . . . .  N. J. Cox
        Q1/10   SJ 10(1):164
        new reverse(ge) option specifies plotting probabilities or
        frequencies greater than or equal to any data value

SJ-5-3  gr0018  . . . . . . . . . .  Speaking Stata: The protean quantile plot
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q3/05   SJ 5(3):442--460           (see gr41_3 and gr42_3 for commands)
        discusses quantile and distribution plots as used in
        the analysis of species abundance data in ecology

SJ-5-3  gr41_3  . . . . . . . . . . . . . . . . . Software update for distplot
        (help distplot if installed)  . . . . . . . . . . . . . . .  N. J. Cox
        Q3/05   SJ 5(3):471
        simplified syntax; both by() and over() are now allowed

SJ-4-2  gr0004  .  Speaking Stata: Graphing categorical and compositional data
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q2/04   SJ 4(2):190--215                                 (no commands)
        discusses graphical possibilities for categorical and
        compositional data

SJ-4-1  gr0003  . . . . . . . . . . . . Speaking Stata: Graphing distributions
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q1/04   SJ 4(1):66--88                                   (no commands)
        a review of official and user-written commands for
        graphing univariate distributions; includes tricks
        beyond what is obviously and readily available

SJ-3-4  gr41_2  . . . . . . . . . . . . . . . . . Software update for distplot
        (help distplot if installed)  . . . . . . . . . . . . . . .  N. J. Cox
        Q4/03   SJ 3(4):449
        option tscale() renamed as trscale()

SJ-3-2  gr41_1  . . . . . . . . . . . . . . . . . Software update for distplot
        (help distplot if installed)  . . . . . . . . . . . . . . .  N. J. Cox
        Q2/03   SJ 3(2):211
        enhanced to use Stata 8 graphics and provides new options

STB-51  gr41  . . . . . . . . . . . . . . . . . .  Distribution function plots
        (help distplot if installed)  . . . . . . . . . . . . . . .  N. J. Cox
        9/99    pp.12--16; STB Reprints Vol 9, pp.108--112
        plots the cumulative distribution function or survival function
        and allows multiple variables

Here is some sample code for your data example. Some of the option choices, notably

xsc(alt)

, are just for fun.

distplot placementdays1, trscale(100 * @) reverse ytitle(% still in care) yla(, ang(h)) xtitle(Placement days) xsc(alt)

[Nick Cox]

Here's the graph again if you can't see it easily.

[Collin James]

Thank you both. Both of these solutions work---trying to decide what to use, but I appreciate it.

[Collin James]

Nick, on distplot, is there anyway to show only the graph from 30 days on? I tried dropping and also specifinyg if placementdays1>=30 but it will still show from 0 to the end of placementdays

[Collin James]

I tried xscale(range(30 100)) but didn't seem to work

[Hemanshu Kumar]

Collin James I want to flag an important difference that arises when you look at a subset of the placementdays1 domain. Let me do this using a much larger number than 30, so it's easy to see.

If you do

distplot placementdays1 if placementdays1>=500, trscale(100 * @) reverse ytitle(% still in care) yla(, ang(h)) xtitle(Placement days) xsc(alt)

the code treats the values greater 500 as the entire domain, and computes percentages based on that. The resulting graph looks like this:

On the other hand, if you use the solution in #3 and introduce the if condition only at the graphing stage, i.e.

contract placementdays1, cpercent(cumulperc)
gen float still_in_care = 100 - cumulperc

twoway line still_in_care placementdays1 if placementdays1>=500, xtitle("Placement days") ytitle("% still in care")

then the percentages are computed over the entire domain (from 0 onwards) and the plot is restricted to the domain above 500. So you get:

While for obvious reasons, the shape of the two graphs is similar, notice the y-axis: the percentages being graphed are very different. My sense is you want the latter, not the former.

[Nick Cox]

Hemanshu Kumar is bang on. An if qualifier with distplot doesn't calculate with all of the data but select some of the results to show; it restricts the calculation to the data selected. That is pretty much standard Stata behaviour.

Also range() within yscale() or xscale() never selects part of the graph; it is not a zooming suboption. This is documented within the help for axis scale options.

range() never narrows the scale of an axis or causes data to be omitted from the plot.

What you want naturally makes perfect sense but #7 and #8 don't work in that way. Hemanshu explained how to get it.

Note also cumul as an official command long predating distplot.

[Collin James]

Thank you both---there is one issue with the solution in #3. I want to plot two different lines by a variable called inc_parent (that takes values 1 or 0). Do I have to include this in the contract statement in some way? And if so, what does it do when it contracts my variable "inc_parent." Yes, I see what you mean about how the distribution is calculated. This is very helpful. Thanks!

[Nick Cox]

The best way to understand it is to try it. I would contract to frequencies first, and then calculate what is wanted afterwards. See also cumul as flagged in #7,