1 Time-stamp: <2006-02-17 22:06:31 blp>
3 Get rid of need for GNU diff in `make check'.
5 Get rid of need for file name canonicalization.
7 Format specifier code needs to be rewritten for lowered crappiness.
9 CROSSTABS needs to be re-examined.
11 RANK, which is needed for the Wilcoxon signed-rank statistic, Mann-Whitney U,
12 Kruskal-Wallis on NPAR TESTS and for Spearman and the Johnkheere trend test (in
15 lex_token_representation() should take a buffer to fill.
17 Make valgrind --leak-check=yes --show-reachable=yes work.
19 Add NOT_REACHED() macro.
21 Add compression to casefiles.
23 Scratch variables should not be available for use following TEMPORARY.
25 Check our results against the NIST StRD benchmark results at
26 strd.itl.nist.gov/div898/strd
28 In debug mode hash table code should verify that collisions are reasonably low.
30 Use AFM files instead of Groff font files, and include AFMs for our default
31 fonts with the distribution.
33 Storage of value labels on disk is inefficient. Invent new data structure.
35 Add an output flag which would cause a page break if a table segment could fit
36 vertically on a page but it just happens to be positioned such that it won't.
38 Fix spanned joint cells, i.e., EDLEVEL on crosstabs.stat.
42 PostScript driver should emit thin lines, then thick lines, to optimize time
45 Should be able to bottom-justify cells. It'll be expensive, though, by
46 requiring an extra metrics call.
48 Perhaps instead of the current lines we should define the following line types:
49 null, thin, thick, double. It might look pretty classy.
51 Perhaps thick table borders that are cut off by a page break should decay to
52 thin borders. (i.e., on a thick bordered table that's longer than one page,
53 but narrow, the bottom border would be thin on the first page, and the top and
54 bottom borders on middle pages.)
56 Support multi-line titles on tables. (For the first page only, presumably.)
58 In addition to searching the source directory, we should search the current
59 directory (for data files). (Yuck!)
61 Fix line-too-long problems in PostScript code, instead of covering them up.
62 setlinecap is *not* a proper solution.
66 Improve interactivity of output by allowing a `commit' function for a page.
67 This will also allow for infinite-length pages.
69 Implement thin single lines, should be pretty easy now.
71 SELECT IF should be moved before other transformations whenever possible. It
72 should only be impossible when one of the variables referred to in SELECT IF is
73 created or modified by a previous transformation.
75 The manual: add text, add index entries, add examples.
77 Figure out a stylesheet for messages displayed by PSPP: i.e., what quotation
78 marks around filenames, etc.
80 From Zvi Grauer <z.grauer@csuohio.edu> and <zvi@mail.ohio.net>:
82 1. design of experiments software, specifically Factorial, response surface
83 methodology and mixrture design.
85 These would be EXTREMELY USEFUL for chemists, engineeris, and anyone
86 involved in the production of chemicals or formulations.
88 2. Multidimensional Scaling analysis (for market analysis) -
90 3. Preference mapping software for market analysis
92 4. Hierarchical clustering (as well as partition clustering)
96 6. Categorical data analsys ?
101 Sometimes very wide (or very tall) columns can occur in tables. What is a good
102 way to truncate them? It doesn't seem to cause problems for the ascii or
103 postscript drivers, but it's not good in the general case. Should they be
104 split somehow? (One way that wide columns can occur is through user request,
105 for instance through a wide PRINT request--try time-date.stat with a narrow
106 ascii page or with the postscript driver on letter size paper.)
108 NULs in input files break the products we're replacing: although it will input
109 them properly and display them properly as AHEX format, it truncates them in A
110 format. Also, string-manipulation functions such as CONCAT truncate their
111 results after the first NUL. This should simplify the result of PSPP design.
112 Perhaps those ugly a_string, b_string, ..., can all be eliminated.
114 From Moshe Braner <mbraner@nessie.vdh.state.vt.us>: An idea regarding MATCH
115 FILES, again getting BEYOND the state of SPSS: it always bothered me that if I
116 have a large data file and I want to match it to a small lookup table, via
117 MATCH FILES FILE= /TABLE= /BY key, I need to SORT the large file on key, do the
118 match, then (usually) re-sort back into the order I really want it. There is
119 no reason to do this, when the lookup table is small. Even a dumb sequential
120 search through the table, for every case in the big file, is better, in some
121 cases, than the sort. So here's my idea: first look at the /TABLE file, if it
122 is "small enough", read it into memory, and create an index (or hash table,
123 whatever) for it. Then read the /FILE and use the index to match to each case.
124 OTOH, if the /TABLE is too large, then do it the old way, complaining if either
125 file is not sorted on key.
127 ----------------------------------------------------------------------
128 Statistical procedures:
130 For each case we read from the input program:
132 1. Execute permanent transformations. If these drop the case, stop.
133 2. N OF CASES. If we have already written N cases, stop.
134 3. Write case to replacement active file.
135 4. Execute temporary transformations. If these drop the case, stop.
136 5. Post-TEMPORARY N OF CASES. If we have already analyzed N cases, stop.
137 6. FILTER, PROCESS IF. If these drop the case, stop.
138 7. Pass case to procedure.
142 LAG records cases in step 3.
144 AGGREGATE: When output goes to an external file, this is just an ordinary
145 procedure. When output goes to the active file, step 3 should be skipped,
146 because AGGREGATE creates its own case sink and writes to it in step 7. Also,
147 TEMPORARY has no effect and we just cancel it. Regardless of direction of
148 output, we should not implement AGGREGATE through a transformation because that
149 will fail to honor FILTER, PROCESS IF, N OF CASES.
151 ADD FILES: Essentially an input program. It silently cancels unclosed LOOPs
152 and DO IFs. If the active file is used for input, then runs EXECUTE (if there
153 are any transformations) and then steals vfm_source and encapsulates it. If
154 the active file is not used for input, then it cancels all the transformations
155 and deletes the original active file.
161 MATCH FILES: Similar to AGGREGATE. This is a procedure. When the active file
162 is used for input, it reads the active file; otherwise, it just cancels all the
163 transformations and deletes the original active file. Step 3 should be
164 skipped, because MATCH FILES creates its own case sink and writes to it in step
165 7. TEMPORARY is not allowed.
173 UPDATE: same as ADD FILES.
176 ----------------------------------------------------------------------
179 * Before TEMPORARY, limits number of cases sent to the sink.
181 * After TEMPORARY, limits number of cases sent to the procedure.
183 * Without TEMPORARY, those are the same cases, so it limits both.
187 * Sample is just a transformation. It has no special properties.
191 * Always selects cases sent to the procedure.
193 * No effect on cases sent to sink.
195 * Before TEMPORARY, selection is permanent. After TEMPORARY,
196 selection stops after a procedure.
200 * Always selects cases sent to the procedure.
202 * No effect on cases sent to sink.
204 * Always stops after a procedure.
208 * Ignored by AGGREGATE. Used when procedures write matrices.
210 * Always applies to the procedure.
212 * Before TEMPORARY, splitting is permanent. After TEMPORARY,
213 splitting stops after a procedure.
217 * TEMPORARY has no effect on AGGREGATE when output goes to the active file.
219 * SORT CASES, ADD FILES, RENAME VARIABLES, CASESTOVARS, VARSTOCASES,
220 COMPUTE with a lag function cannot be used after TEMPORARY.
222 * Cannot be used in DO IF...END IF or LOOP...END LOOP.
224 * FLIP ignores TEMPORARY. All transformations become permanent.
226 * MATCH FILES and UPDATE cannot be used after TEMPORARY if active
227 file is an input source.
229 * RENAME VARIABLES is invalid after TEMPORARY.
231 * WEIGHT, SPLIT FILE, N OF CASES, FILTER, PROCESS IF apply only to
232 the next procedure when used after TEMPORARY.
236 * Always applies to the procedure.
238 * Before TEMPORARY, weighting is permanent. After TEMPORARY,
239 weighting stops after a procedure.
242 -------------------------------------------------------------------------------