1 Time-stamp: <2005-03-02 16:08:59 blp>
3 What Ben's working on now.
4 --------------------------
6 Workspace exhaustion heuristics.
8 Update q2c input format description.
10 Rewrite output subsystem, break into multiple processes.
12 CROSSTABS needs to be re-examined.
14 RANK, which is needed for the Wilcoxon signed-rank statistic, Mann-Whitney U,
15 Kruskal-Wallis on NPAR TESTS and for Spearman and the Johnkheere trend test (in
21 Make valgrind --leak-check=yes --show-reachable=yes work.
23 Add NOT_REACHED() macro.
25 Add compression to casefiles.
27 There needs to be another layer onto the lexer, which should probably be
28 entirely rewritten anyway. The lexer needs to read entire *commands* at a
29 time, not just a *line* at a time. It also needs to support arbitrary putback,
30 probably by just backing up the "current position" in the command buffer.
32 Scratch variables should not be available for use following TEMPORARY.
34 Details of N OF CASES, SAMPLE, FILTER, PROCESS IF, TEMPORARY, etc., need to be
35 checked against the documentation. See notes on these at end of file for a
38 Check our results against the NIST StRD benchmark results at
39 strd.itl.nist.gov/div898/strd
41 In debug mode hash table code should verify that collisions are reasonably low.
43 Use AFM files instead of Groff font files, and include AFMs for our default
44 fonts with the distribution.
46 Storage of value labels on disk is inefficient. Invent new data structure.
48 Add an output flag which would cause a page break if a table segment could fit
49 vertically on a page but it just happens to be positioned such that it won't.
51 Fix spanned joint cells, i.e., EDLEVEL on crosstabs.stat.
55 PostScript driver should emit thin lines, then thick lines, to optimize time
58 New functions? var_name_or_label(), tab_value_or_label()
60 Should be able to bottom-justify cells. It'll be expensive, though, by
61 requiring an extra metrics call.
63 Perhaps instead of the current lines we should define the following line types:
64 null, thin, thick, double. It might look pretty classy.
66 Perhaps thick table borders that are cut off by a page break should decay to
67 thin borders. (i.e., on a thick bordered table that's longer than one page,
68 but narrow, the bottom border would be thin on the first page, and the top and
69 bottom borders on middle pages.)
71 Support multi-line titles on tables. (For the first page only, presumably.)
73 Rewrite the convert_F() function in data-out.c to be nicer code.
75 In addition to searching the source directory, we should search the current
76 directory (for data files). (Yuck!)
78 Fix line-too-long problems in PostScript code, instead of covering them up.
79 setlinecap is *not* a proper solution.
83 Has glob.c been pared down enough?
85 Improve interactivity of output by allowing a `commit' function for a page.
86 This will also allow for infinite-length pages.
88 Implement thin single lines, should be pretty easy now.
90 SELECT IF should be moved before other transformations whenever possible. It
91 should only be impossible when one of the variables referred to in SELECT IF is
92 created or modified by a previous transformation.
94 The manual: add text, add index entries, add examples.
96 The inline file should be improved: There should be *real* detection of whether
97 it is used (in dfm.c:cmd_begin_data), not after-the-fact detection.
99 Figure out a stylesheet for messages displayed by PSPP: i.e., what quotation
100 marks around filenames, etc.
102 New SET subcommand: OUTPUT. i.e., SET OUTPUT="filename" to send output to that
103 file; SET OUTPUT="filename"(APPEND) to append to that file; SET OUTPUT=DEFAULT
104 to reset everything. There might be a better approach, though--think about it.
106 From Zvi Grauer <z.grauer@csuohio.edu> and <zvi@mail.ohio.net>:
108 1. design of experiments software, specifically Factorial, response surface
109 methodology and mixrture design.
111 These would be EXTREMELY USEFUL for chemists, engineeris, and anyone
112 involved in the production of chemicals or formulations.
114 2. Multidimensional Scaling analysis (for market analysis) -
116 3. Preference mapping software for market analysis
118 4. Hierarchical clustering (as well as partition clustering)
122 6. Categorical data analsys ?
127 In addition to an "infinite journal", we should keep a number of
128 individual-session journals, pspp.jnl-1 through pspp.jnl-X, renaming and
129 deleting as needed. All of the journals should have date/time comments.
131 Qualifiers for variables giving type--categorical, ordinal, ...
135 Consider consequences of xmalloc(), fail(), hcf() in interactive
137 a. Can we safely just use setjmp()/longjmp()?
138 b. Will that leak memory?
139 i. I don't think so: all procedure-created memory is either
140 garbage-collected or globally-accessible.
141 ii. But you never know... esp. w/o Checker.
142 c. Is this too early to worry? too late?
144 Need to implement a shared buffer for funny functions that require relatively
145 large permanent transient buffers (1024 bytes or so), that is, buffers that are
146 permanent in the sense that they probably shouldn't be deallocated but are only
147 used from time to time, buffers that can't be allocated on the stack because
148 they are of variable and unpredictable but usually relatively small (usually
149 line buffers). There are too many of these lurking around; can save a sizeable
150 amount of space at very little overhead and with very little effort by merging
153 Clever multiplatform GUI idea (due partly to John Williams): write a GUI in
154 Java where each statistical procedure dialog box could be downloaded from the
155 server independently. The statistical procedures would run on (the/a) server
156 and results would be reported through HTML tables viewed with the user's choice
157 of web browsers. Help could be implemented through the browser as well.
162 MORE NOTES/IDEAS/BUGS
163 ---------------------
165 Sometimes very wide (or very tall) columns can occur in tables. What is a good
166 way to truncate them? It doesn't seem to cause problems for the ascii or
167 postscript drivers, but it's not good in the general case. Should they be
168 split somehow? (One way that wide columns can occur is through user request,
169 for instance through a wide PRINT request--try time-date.stat with a narrow
170 ascii page or with the postscript driver on letter size paper.)
172 NULs in input files break the products we're replacing: although it will input
173 them properly and display them properly as AHEX format, it truncates them in A
174 format. Also, string-manipulation functions such as CONCAT truncate their
175 results after the first NUL. This should simplify the result of PSPP design.
176 Perhaps those ugly a_string, b_string, ..., can all be eliminated.
178 From Moshe Braner <mbraner@nessie.vdh.state.vt.us>: An idea regarding MATCH
179 FILES, again getting BEYOND the state of SPSS: it always bothered me that if I
180 have a large data file and I want to match it to a small lookup table, via
181 MATCH FILES FILE= /TABLE= /BY key, I need to SORT the large file on key, do the
182 match, then (usually) re-sort back into the order I really want it. There is
183 no reason to do this, when the lookup table is small. Even a dumb sequential
184 search through the table, for every case in the big file, is better, in some
185 cases, than the sort. So here's my idea: first look at the /TABLE file, if it
186 is "small enough", read it into memory, and create an index (or hash table,
187 whatever) for it. Then read the /FILE and use the index to match to each case.
188 OTOH, if the /TABLE is too large, then do it the old way, complaining if either
189 file is not sorted on key.
191 ----------------------------------------------------------------------
192 Statistical procedures:
194 For each case we read from the input program:
196 1. Execute permanent transformations. If these drop the case, stop.
197 2. N OF CASES. If we have already written N cases, stop.
198 3. Write case to replacement active file.
199 4. Execute temporary transformations. If these drop the case, stop.
200 5. Post-TEMPORARY N OF CASES. If we have already analyzed N cases, stop.
201 6. FILTER, PROCESS IF. If these drop the case, stop.
202 7. Pass case to procedure.
206 LAG records cases in step 3.
208 AGGREGATE: When output goes to an external file, this is just an ordinary
209 procedure. When output goes to the active file, step 3 should be skipped,
210 because AGGREGATE creates its own case sink and writes to it in step 7. Also,
211 TEMPORARY has no effect and we just cancel it. Regardless of direction of
212 output, we should not implement AGGREGATE through a transformation because that
213 will fail to honor FILTER, PROCESS IF, N OF CASES.
215 ADD FILES: Essentially an input program. It silently cancels unclosed LOOPs
216 and DO IFs. If the active file is used for input, then runs EXECUTE (if there
217 are any transformations) and then steals vfm_source and encapsulates it. If
218 the active file is not used for input, then it cancels all the transformations
219 and deletes the original active file.
225 MATCH FILES: Similar to AGGREGATE. This is a procedure. When the active file
226 is used for input, it reads the active file; otherwise, it just cancels all the
227 transformations and deletes the original active file. Step 3 should be
228 skipped, because MATCH FILES creates its own case sink and writes to it in step
229 7. TEMPORARY is not allowed.
237 UPDATE: same as ADD FILES.
240 ----------------------------------------------------------------------
243 * Before TEMPORARY, limits number of cases sent to the sink.
245 * After TEMPORARY, limits number of cases sent to the procedure.
247 * Without TEMPORARY, those are the same cases, so it limits both.
251 * Sample is just a transformation. It has no special properties.
255 * Always selects cases sent to the procedure.
257 * No effect on cases sent to sink.
259 * Before TEMPORARY, selection is permanent. After TEMPORARY,
260 selection stops after a procedure.
264 * Always selects cases sent to the procedure.
266 * No effect on cases sent to sink.
268 * Always stops after a procedure.
272 * Ignored by AGGREGATE. Used when procedures write matrices.
274 * Always applies to the procedure.
276 * Before TEMPORARY, splitting is permanent. After TEMPORARY,
277 splitting stops after a procedure.
281 * TEMPORARY has no effect on AGGREGATE when output goes to the active file.
283 * SORT CASES, ADD FILES, RENAME VARIABLES, CASESTOVARS, VARSTOCASES,
284 COMPUTE with a lag function cannot be used after TEMPORARY.
286 * Cannot be used in DO IF...END IF or LOOP...END LOOP.
288 * FLIP ignores TEMPORARY. All transformations become permanent.
290 * MATCH FILES and UPDATE cannot be used after TEMPORARY if active
291 file is an input source.
293 * RENAME VARIABLES is invalid after TEMPORARY.
295 * WEIGHT, SPLIT FILE, N OF CASES, FILTER, PROCESS IF apply only to
296 the next procedure when used after TEMPORARY.
300 * Always applies to the procedure.
302 * Before TEMPORARY, weighting is permanent. After TEMPORARY,
303 weighting stops after a procedure.
306 -------------------------------------------------------------------------------