File: CHANGES_FROM_ESL

package info (click to toggle)
gretl 0.99.2-1
  • links: PTS
  • area: main
  • in suites: woody
  • size: 7,304 kB
  • ctags: 3,210
  • sloc: ansic: 43,338; sh: 6,567; makefile: 807; perl: 529
file content (111 lines) | stat: -rw-r--r-- 4,229 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
Here is an accounting of the changes from Ramu Ramanathan's
"ESL", in my "esclient".

1. Change of syntax
2. Change in command-line arguments
3. Commands missing from eslclient
4. Commands redefined in eslclient
5. New commands added to eslclient
6. Other features added to eslclient


1. Change of syntax
===================

There is only one significant change.  In esl, a semicolon is used as a
terminator for many commands.  I decided to remove this in eslclient.
Semicolons are simply ignored, apart from a few special cases where
they have a definite meaning: as a separator for two lists in the "ar"
and "tsls" commands, and as a marker for an unchanged starting or ending
obesrvation in the "smpl" command.

In esl semicolon termination gives the possibility of breaking long
commands over more than one line; in eslclient this is done by putting a
trailing backslash "\" at the end of a line that is to be continued.

2. Change in command-line arguments
===================================

The command-line syntax for running a batch job is simplified.  For esl
you type, e.g.

  esl -b datafile <inputfile >outputfile
  
For eslclient you type:

  eslclient -b inputfile >outputfile  

The inputfile is treated as a program argument; it should specify a
datafile to use internally, using 

  open datafile  OR  (* ! datafile *)
  
3. Commands missing from eslclient
==================================

I have not implemented the esl commands designed to make working
interactively at the DOS command prompt a bit easier ("scroll" and
"edit").  I presume that with the new gui these will not be needed, and
that people who choose to use the command-line interface interactively
will probably be running it in a proper scrollable terminal window (e.g.
xterm).

4. Commands redefined in eslclient
==================================

A few esl commands have been simplified, or their output has been
changed somewhat.

freq: At present you can't specify particular ranges as in esl.  A
chi-square test for normality has been added.

store: The -o switch now saves the data by variable in binary format. 
There are two new switches: -c saves in CSV format; -m saves in GNU
Octave format.

The output from many commands is formatted a little differently.

5. New commands added to eslclient
==================================

These are described in detail in the helpfile.  Here is a brief summary.

adf:		Augmented Dickey-Fuller test
chow:		Chow test for structural break
coint:		Cointegration test
diff:		Shortcut for generating first differences
endloop:    Terminate a monte carlo simulation loop
eqnprint:	Print OLS model as LaTeX equation
import: 	Open datafile in comma-separated values (CSV) format
info:       Display any info on data in header file		
labels:     Display explanatory labels for vars (if any)
ldiff:		Shortcut for generating log-differences
loop:		Start a monte carlo simulation loop
lsqprint:	Print OLS model as LaTeX table
nulldata:	Set up blank "dummy" data set (for simulations)
open: 		Open an esl format datafile

6. Other features added to eslclient
====================================

* Specifying lags: You don't have to create lagged variables in advance
of a regression.  The syntax "foo(-1)" in a regression list will cause
the first lag of foo (if foo exists) to be generated and added to the
data set.

* You can switch data files without quitting eslclient.  If a data file
has already been read, using the "open" command will replace it with a
new one (after a prompt for whether this is really wanted).

* You can assign "labels" to variables: explanatory or descriptive tags
of up to 128 characters.  These go in a file with the same basename as
the datafile plus the suffix ".lbl".  One variable per line, the name of
the variable first, followed by white space, followed by the label. 
These labels will be read if they're present and can be retrieved with
the "labels" command.

* case markers: The first column of a data file may contain case markers
(8 chars max), such as the names of states or cities.  This must be
flagged by the keyword MARKERS at the end of the data header file.  When
data are printed, the case markers will be used instead of observation
numbers.