[soar] r14068 committed - Edited wiki page cmd_smem through web user interface.

[so...@googlecode.com]

Revision: 14068
Author: maz...@gmail.com
Date: Wed Sep 10 21:57:32 2014 UTC
Log: Edited wiki page cmd_smem through web user interface.
https://code.google.com/p/soar/source/detail?r=14068

Modified:
/wiki/cmd_smem.wiki

=======================================
--- /wiki/cmd_smem.wiki Wed Jun 25 00:52:16 2014 UTC
+++ /wiki/cmd_smem.wiki Wed Sep 10 21:57:32 2014 UTC
@@ -23,6 +23,7 @@

== Options ==

+|| *Option* || *Description* ||
|| `-e, --enable, --on` || Enable semantic memory. ||
|| `-d, --disable, --off` || Disable semantic memory. ||
|| `-g, --get` || Print current parameter setting ||
@@ -31,7 +32,10 @@
|| `-S, --stats` || Print statistic summary or specific statistic ||
|| `-t, --timers` || Print timer summary or specific statistic ||
|| `-a, --add` || Add concepts to semantic memory ||
+|| `-r, --remove`|| Remove concepts from semantic memory||
|| `-p, --print` || Print semantic store in user-readable format ||
+|| `-q, --query`|| Print concepts in semantic store matching some cue||
+|| `-h, --history`|| Print activation history for some LTI||
|| `-v, --viz` || Print semantic store in graphviz format ||
|| `-b, --backup` || Creates a backup of the semantic database on disk ||

@@ -168,10 +172,12 @@

Level one

+|| *Timer* || *Description* ||
|| `_total` || Total smem operations ||

Level two

+|| *Timer* || *Description* ||
|| `smem_api` || Agent command validation ||
|| `smem_hash` || Hashing symbols ||
|| `smem_init` || Semantic store initialization ||
@@ -181,6 +187,7 @@

Level three

+|| *Timer* || *Description* ||
|| three_activation || Recency information maintenance ||

=== Manual Storage ===
@@ -206,6 +213,69 @@
"arithmetic" variable) with three children. Each child will be its
own concept with two constant attribute/value pairs.

+=== Manual Removal ===
+
+Part or all of the information in the semantic store of some LTI can be
manually removed from the semantic store using the
+
+{{{
+smem --remove <concept>
+}}}
+
+command. The format for specifying what to remove is similar to that of
+adding WMEs to working memory on the RHS of productions. For example:
+{{{
+smem --remove {
+ (@W34 ^good-attribute |gibberish value|)
+}
+}}}
+If `good-attribute` is multi-valued, then all values will remain in the
store except `|gibberish value|`. If `|gibberish value|` is the only
value, then `good-attribute` will also be removed. It is not possible to
use the common "dot-notation" for this command. Manual removal is not
recursive.
+
+Another example highlights the ability to remove all of the values for an
attribute:
+{{{
+smem --remove {
+ (@W34 ^bad-attribute)
+}
+}}}
+When a value is not given, all of the values for the given attribute are
removed from the LTI in the semantic store.
+
+Also, it is possible to remove all augmentations of some LTI from the
semantic store:
+{{{
+smem --remove {
+ (@W34)
+}
+}}}
+This would remove all attributes and values of `@W34` from the semantic
store. The LTI will remain in the store, but will lack augmentations.
+
+(Use the following at your own risk.) Optionally, the user can force
removal even in the event of an error:
+{{{
+smem -r {(@W34 ^bad-attribute ^bad-attribute-2)} force
+}}}
+Suppose that LTI `@W34` did not contain `bad-attribute`. The above
example would remove `bad-attribute-2` even though it would indicate an
error (having not found `bad-attribute`).
+
+=== Manual Query ===
+
+Queries for LTIs in the semantic store that match some cue can be
initialized external to an agent using the
+`smem --query <cue> [<num>]`
+command. The format for specifying the cue is similar to that of adding a
new identifier to working memory in the RHS of a rule:
+{{{
+smem --query {
+ (<cue> ^attribute <wildcard> ^attribute-2 |constant|)
+}
+}}}
+Note that the root of the cue structure must be a variable and should be
unused in the rest of the cue structure. This command is for testing and
the full range of queries accessible to the agent are not yet available for
the command. For example, math queries are not supported.
+
+The additional option of `<num>` will trigger the display of the top
`<num>` most activated LTIs that matched the cue.
+
+The result of a manual query is either to print that no LTIs could be
found or to print the information associated with LTIs that were found in
the `smem --print <lti>` format.
+
+=== History ===
+
+When the activation-mode of a semantic store is set to base-level, some
history of activation events is stored for each LTI. This history of when
some LTI was activated can be displayed:
+{{{
+ smem --history @W34
+}}}
+In the event that semantic memory is not using base-level activation,
`history` will mimic `print`.
+
=== Visualization ===

When debugging agents using semantic memory it is often useful to