Stierhof: copy from old wiki

2018-04-11T12:04:41Z

copy from old wiki

New page

=== How can I find out which function/line in my code is causing a stack overflow? ===
You can use the package '''''stkcheck''''' and activate the function '''''enable_stack_check()''''':
<pre>
require("stkcheck");
enable_stack_check();

printf("Guess how many characters were written...\n");
</pre>

This will show in which line/functions values are left on the stack:
<pre>
Guess how many characters were written...
<stdin>:4: 1 object(s) left on the stack
42
</pre>

You have to know that in S-Lang, you have to explicitly catch all return values of the functions you're calling.
In case you're not interested in the return value, discard it by assigning it to an empty pair of parentheses:
<pre>
()=printf("I don't care about the number of bytes written!\n");
</pre>

<div class="mw-collapsible mw-collapsed" data-expandtext="show" data-collapsetext="hide">
====== More information for experts ======
----
<div class="mw-collapsible-content">
S-Lang functions may have more than one return value. A trivial example:
<pre>
define sum_and_product(a, b)
{
return (a+b, a*b);
}

(s, p) = sum_and_product(a0, b0);
</pre>

One selectively discard ''some'' return values, or also ''all of them'',
by using according empty places in the comma-seaparated list of variables inside the parentheses:
<pre>
(s, ) = sum_and_product(a0, b0);
( , p) = sum_and_product(a0, b0);
( , ) = sum_and_product(a0, b0);
</pre>
</div></div>

Note that in versions of S-Lang prior to [http://www.sternwarte.uni-erlangen.de/~hanke/git.slang/?p=S-Lang;a=commit;h=11dab02ad2f43a22337f8a06d9e2938ceead03a4 pre2.2.4-7],
qualifier values were confused with values left on the stack,
such that <tt>enable_stack_check()</tt> would complain in any case.

=== How can I read CSV files? ===
CSV files are Comma-Seperated Variable ASCII files, which are easily read into ISIS using the <tt>ascii_read_table</tt> function. For each value in your file you have to give the format plus the separator to the next value, i.e., a comma (",").
<pre>
variable dat = ascii_read_table("lc.csv", [{"%F,","time"},{"%F,","rate"},{"%F","err"}]) ;
</pre>
This returns a structure <tt>dat</tt>
<pre>
dat = struct{Double_Type time, Double_Type rate, Double_Type err}
</pre>

or you can use the presumably better <tt>csv</tt> routines in ISIS, which you can get access to by <tt>require("csv")</tt>. You can then read your data using
<pre>
reqiuire("csv") ;
variable dat = csv_readcol ("lc.csv" ; type=['d','d','d'], fields = ["time", "rate", "err"]) ;
</pre>
which returns the same structure as above. Note that <tt>type</tt> needs to be an array of <tt>UChar</tt>s, so single quotes only. Without the <tt>type</tt> qualifier the fields of <tt>dat</tt> will be <tt>String_Type</tt>.

See the documentation of <tt>csv_readcol</tt> and <tt>csv.readcol</tt> for further information.

[[Category:Isis / Slang]]

Questions and Answers - Revision history

Stierhof: copy from old wiki