Dear Jens
A request. Could you or any of the resource people make a short podcast on the use of the updated epidata analysis?
Or are there any resources available for this online i/ we can access
Thanks in advance
Thambu
Thank you for your response.
Example with my own file.
.freq castem;
castem
-----------
N
-----------
1 153
2 150
3 632
-----------
Total 935
-----------
.select castem<>3 do;
Selecting, please wait...
(303 of 935 selected)
.begin; ==> No operative (idem begin without <
; >)
ERROR: Line 1: Syntax error at pos 7:
ERROR: Unexpected end of line!
Perhaps a ";" is missing?
.freq castem; ==> no selection...
castem
-----------
N
-----------
1 153
2 150
3 632
-----------
Total 935
-----------
.select castem=1 do drop data; ==> objective : keep codage 1
Selecting, please wait...
(153 of 935 selected)
(153 observations dropped)
.freq castem; ==> Selection code<> 1 (link avec drop data ?)
castem
-----------
N
-----------
2 150
3 632
-----------
Total 782
-----------
And an another (and last) question : how to delete the current selection?
Thank you.
Bernard Branger, Nantes, France
Hello,
For training, I'm trying to understand the new version of "Analysis".
A few comments :
- The command "freq var ! r;" (as shown in the guide) doesn't work, but
"freq var ! pr" is OK.
Strictly speaking and on the educational plan, the percentages presented
with "freq" are in columns : "freq var ! c;" or "freq var ! pc;" would be
more logical.
- To select a part of the file, I noted, (as an example for a frequency) the
command "select (var1 <3) do freq var2 ! pr;".
Should we repeat with each command "select (var1 <3) do"?
Is there a way, like the old "select", to select part of the file for
further commands?
- The command (example) "select (var1 = 3) do var1: = 2;" is quite
complicated compared to that "if ... then ...."
- How to create a new numeric variable with 3 digits and 2 decimal places?
As in the old "define var ###. ##"?
- When will the "regress" commands and the graph commands (including scatter
with a regression line as in the old "Epi-Info" ...?
Thank you for your help.
Bernard BRANGER - 44000 NANTES - France
