由买买提看人间百态

boards

本页内容为未名空间相应帖子的节选和存档,一周内的贴子最多显示50字,超过一周显示500字 访问原贴
Statistics版 - SAS Graphic question - help please
相关主题
请教SAS LABEL问题。Ask 2 simple SAS questions,thanks
SAS问题来了再来问一个SAS问题
SAS问题A SAS question
请问R Code和 Matlab Code 思路是否近似?SAS question - baozi
请教一个SAS小问题:如何得到 one cumulative varialbe of character variable?SAS question - baozi
SAS question: count non-missing value for different variables急问一个call symput问题(SAS)
[合集] how to calculate column sum not row sum in SAS? thanks a lo一个SAS的时间区域求值问题
SAS 请教双包子求教:SAS问题
相关话题的讨论汇总
话题: sas话题: 12am话题: graphic话题: 12pm话题: 86400
进入Statistics版参与讨论
1 (共1页)
y*****w
发帖数: 1350
1
In SAS, how to make the X-axis labels as in the attached graph? The raw data
I have is per minute (or the clock time like 12:00:00) along days. Thanks!
w*******n
发帖数: 469
2
format
s*****a
发帖数: 2735
3
axis1 label=none order=(1 to 13)/*??x axis 的范围*/
value=(j=c "12AM" " " "12PM" " ""12AM" " ""12PM" " ""12AM"
" ""12PM" " ""12AM");
/* ..... */
proc gplot;
plot var1*var2 / haxis = axis1;
run;
/* ???可行 *//
y*****w
发帖数: 1350
4
Thanks for the above responses. I have worked it out in the AXIS statement
using the same idea as shincha suggested, but the actual data is recorded as
real clock time like 12:04:37 (although every minute), so the actual
numbers in the ORDER option are very large (i.e. 24x3600=86400 per day). By
the way, due to the long days in the actual data, I won't need the minor
ticks.
I also added t=1, t=2, ... (see below) --- is this necessy?
axis1 minor=none label=none
order=(0 to n*86400 by 43200) /* n is the number of days */
value=(t=1 j=c "12AM"
t=2 j=c "12PM"
t=3 j=c "12AM"
t=4 j=c "12PM"
...);
s*****a
发帖数: 2735
5
in my own opinion, "t=xx" is not necessary, unless you need an additional
line for each value label.
I will make a new variable (1 to 13) instead of using "12:04:37" "xxx", (
just my $0.02).

as
By

【在 y*****w 的大作中提到】
: Thanks for the above responses. I have worked it out in the AXIS statement
: using the same idea as shincha suggested, but the actual data is recorded as
: real clock time like 12:04:37 (although every minute), so the actual
: numbers in the ORDER option are very large (i.e. 24x3600=86400 per day). By
: the way, due to the long days in the actual data, I won't need the minor
: ticks.
: I also added t=1, t=2, ... (see below) --- is this necessy?
: axis1 minor=none label=none
: order=(0 to n*86400 by 43200) /* n is the number of days */
: value=(t=1 j=c "12AM"

y*****w
发帖数: 1350
6
I don't understand why it should be (1 to 13). As I mentioned, the data
point can be on any of the 86400 time points on a day.

【在 s*****a 的大作中提到】
: in my own opinion, "t=xx" is not necessary, unless you need an additional
: line for each value label.
: I will make a new variable (1 to 13) instead of using "12:04:37" "xxx", (
: just my $0.02).
:
: as
: By

s*****a
发帖数: 2735
7
sorry, did not read your post carefully.
but why not making a new variable instead? var_new=old_time/86400 (or 43200)?
anyway, it is good to know you fixed the value label.

【在 y*****w 的大作中提到】
: I don't understand why it should be (1 to 13). As I mentioned, the data
: point can be on any of the 86400 time points on a day.

1 (共1页)
进入Statistics版参与讨论
相关主题
双包子求教:SAS问题请教一个SAS小问题:如何得到 one cumulative varialbe of character variable?
弱问一个SAS里面求adjusted means的问题SAS question: count non-missing value for different variables
我也来请教一个SAS问题[合集] how to calculate column sum not row sum in SAS? thanks a lo
如何在R里refer上一个row的值SAS 请教
请教SAS LABEL问题。Ask 2 simple SAS questions,thanks
SAS问题来了再来问一个SAS问题
SAS问题A SAS question
请问R Code和 Matlab Code 思路是否近似?SAS question - baozi
相关话题的讨论汇总
话题: sas话题: 12am话题: graphic话题: 12pm话题: 86400