G7 Commands: L

(le)gend <a> [<b>]

The (le)gend command controls printing a legend at the bottom of a graph.

<a> = y:

for yes, prints the legend (default)

<a> = n:

for no, do not print the legend

<a> = s:

leave space for legend but do not print. It allows the user to annotate the legend.

<b> = y:

for yes, mark the dates (default)

<b> = n:

for no, do not mark dates. It is used to make bar graphs of data occurring at arbitrary intervals.

Related Topics: Drawing graphs in G7 graph

(lgr)aph <name1> [<name2>] [<name3>] … [<name7>] <date1> <date2> [date3]

(lgr)aph (<name1>) [(<name2>)] [(<name3>)] … [(<name7>)] <date1> <date2> [date3]

Construct a Semi-Logarithmic Graph. This is exactly like the graph command except that the series are expected to be in logarithms. The vertical scale will be marked in the natural units, not the units of the logarithms. If vertical range control is in effect (see below), the vertical ranges will be presumed to be in natural units. Algebraic expressions may be provided in place of some or all of the series names, so long as the expressions are enclosed in parentheses. This type of graph also is called a semi-log graph.

Related Topics: graph

(lim)its <date1> <date2> [<date3>]

(lim)its <±n1> <±n2> [<±n3>]

The limits command sets limits for regressions and for @sum() commands. Estimation begins at <date1> and ends at <date2>. Simulations or forecasts will be made to <date3>. Remember that quarterly dates require .1, .2, .3 or .4. Monthly dates require .001 … .012. Annual dates are integer values.

Example:

lim 1975.1 2010.4 2015.3

If valid dates have been specified, then they can be adjusted, moving either or both dates either forward or backward in time.

Example:

lim +0 -2 +1

If the command above follows the command in the first example, then the adjusted limits specify the period quarter one of 1975 through quarter 2 of 2010, and a forecast period through quarter four of 2015.

Related Topics: Dates in G7, Regression, mode, recursive

line <linenumber> <color num | name> <thickness> <mark> <style> <barfill> <left> <right>

line <linenumber>

The line command is used to control the appearance of lines and bars in graphs. The command may be typed interactively, or used in add files to change the appearance of graphs in a “show” file. In G7, the line command can be written from the Graph | Settings dialog box by clicking the Save To File button. This brings up another dialog which asks you for a name of the text file to which you would like to save the graph settings. Suppose you call this file GRAPH.SET. The contents of GRAPH.SET might look something like:

GRAPH.SET
line 1	255	2	+	0	0	0.000	1.000
line 2	16711680	2	s	1	0	0.000	1.000
line 3	32768	2	x	2	0	0.000	1.000
line 4	16776960	2	d	3	0	0.000	1.000
line 5	8388736	2	^	4	0	0.000	1.000
line 6	8388608	2	v	0	0	0.000	1.000
line 7	5315660	2	n	0	0	0.000	1.000

Alternatively, providing the line command followed by a single “linenumber” argument causes the settings for the corresponding line to be printed to screen. Settings for up to 7 series can be specified by the line command. The first argument, <linenumber>, corresponds to the series given in that position on the gr command line. The first series after gr is line 1. The second argument, <color_number>, can be a decimal number representing the components of red, green and blue of the color, or it can be a color name. The third argument, <thickness>, may be set to anything you like, but something in the range of 2 to 4 generally is pleasing. The fourth argument specifies the type of marker on the line, if any. Here is a short table of the possible characters for this argument, and what they mean:

Symbol	Meaning
+	Plus signs
x	X marker
s	Square
d	Diamond
>	Arrow in direction of line
v	Downware pointing triangles
b	Bars
f	Bars with no surrounding rectangle

The next argument is the <style> of the line, possible numeric values for this argument are:

Value	Meaning
0	Solid
1	Dashed
2	Dotted
3	Dash dot
4	Dash dot dot
5	Clear (blank line)

The next argument, <barfill> indicates the type of fill pattern that a bar will take, if a bar graph has been chosen. Possible numeric values for this are:

Value	Meaning
0	Solid fill
1	Clear (empty)
2	Horizontal lines
3	Vertical lines
4	Forward diagonal fill (/////)
5	Backward diagonal fill (\\\\\)
6	Cross hatch
7	Diagonal cross hatch

The <left> and <right> parameters, must be between 0 and 1. They show where, within the space allocated for the bar, the left and right edges of the bar go. For example, with left = 0.1 and right = 0.9, the bar will be in the center of the allowed space, and the bars will be four times as wide (0.8 units) as the space between them (0.2 units). To eliminate the line connecting points marked by bars, set its width to zero. Parallel bars are drawn by assigning non-overlapping intervals to different series. Stacked bars or high-low graphs are produced with overlapping intervals. The table below shows some of the legal color names. Color names may be capitalized or not. Where names have “light” or “dark” these can be abbreviated as “lt” or “dk”. For example, “Lightgray” is the same as “ltgray”.

Basic	Light/Dark	Others
black	lightgray	olive	pumpkin
white	darkgray	lavender	seagreen
red	darkblue	forest	chocolate
green	darkred	aqua	slategray
blue	darkgreen	chartreuse	skyblue
yellow	lightblue	maroon	tomato
orange	lightred	fuchsia	teal
purple	lightgreen	azure	coral
brown	darkbrown	midnight gold

Related Topics: Drawing graphs, Graph settings, vr

lint <name>

lint <group>

Linear interpolation of missing values in a vector is done by the lint command. The lint command replaces missing values in the time series by linearly interpolated values. Zeroes before the first or after the last observation are not replaced. This command applies only to series in the default Vam file. Groups may be used in the <names> field to refer to sectors of the currently loaded vector.

Example:

lint pce1

This example loads the pce vector and then fills in the missing values in sector 1 by linear interpolation. We could similarly fill in the values for all the sectors in the present dynamic group by

load pce
lint :

or all the values for the static group Services by:

load pce
lint :Services

Entire matrices may be interpolated with one command. For example,

lint am

will interpolate the entire am matrix. This interpolation also works for packed matrices. The lint command works on the entire range of the Vam file without regard to fdates.

Related Topics: group

linv [bank letter.]<matrix A> [period]

The linv command calculates the Leontief inverse and places it in the source matrix. It works over the current fdates, or for the [period] specified on the command line.

For example,

linv A 1995

replaces A with its Leontief inverse for 1995.

Related Topics: Matrix operations, madd, mcopy, minv, mmult, mtrans, vc

listbanks (lb)

This function provides information about the banks that are currently assigned in G7. Up to 25 banks of various types may be assigned, to positions ‘a’ through ‘z’ except ‘w’. The lb command reviews which banks are assigned to which position and what type of bank they are.

Related Topics: Assigned banks, bank, cbk, dfr, findmode, hbk, vam

(listg)roups

List the names of all groups currently in the GROUPS.BIN file.

Related Topics: glist, group

(lis)tnames [-srgv] [-l <root>] <bank_location> [wildcard]

Displays the series names in the workspace (w) or the assigned bank (a). For example,

lis a

This command also has a “wildcard” option, which allows you to list variable names matching a certain pattern. For example,

lis a out*

will list all series in the assigned bank that start with out (such as out1 through out85).

Option ‘s’ sorts the series in alphabetical order, and ‘r’ reverses the order. If a Vam file is associated with the G bank, then option ‘g’ prints only macro series and option ‘v’ prints only Vam bank series. If the ‘l’ option is specified and a root name is provided, then the a set of strings will be defined. The string names will be constructed as the root name followed an integer, and the string definition will be the name of the series. One string will be created for each series in the bank.

Related Topics: listnamescol

listnamescol [-srgv] [-l <root>] <bank_location>

lnc [-srgv] [-l <root>] <bank_location>

The listnamescol or lnc command works just like the lis command, but all series names in the workspace (w) or assigned bank (a) are output in one column. You may find it convenient to capture output to a file, and then use this command to get a list of all the series in the databank in that file. Then, using an editor that supports macros, you can change all lines to create an addfile that does the same thing with each series in the databank. Use of wildcard is optional, and like in DOS, ‘*’ will match any number of characters. For any single character, however, you need to use ‘^’, instead of ‘?’ as in DOS. Option ‘s’ sorts the series in alphabetical order, and ‘r’ reverses the order. If a Vam file is associated with the G bank, then option ‘g’ prints only macro series and option ‘v’ prints only Vam bank series. If the ‘l’ option is specified and a root name is provided, then the a set of strings will be defined. The string names will be constructed as the root name followed an integer, and the string definition will be the name of the series. One string will be created for each series in the bank.

Related Topics: bupdate, listnames, save

listvecs

lv [-sr] [<bank_location> [<wildcard>]]

lvc [-sr] [<bank_location> [<wildcard>]]

The listvecs or lv command works just like the lis command, but all vector and matrix names in the default vam file or specified bank are printed. The lnc is the same but prints names in one column. Option ‘s’ sorts the series in alphabetical order, and ‘r’ reverses the order. By default, the command will operate on the default vam bank, but other bank locations may be specified. Use of wildcard is optional, and like in DOS, ‘*’ will match any number of characters. For any single character, however, you need to use ‘^’, instead of ‘?’ as in DOS.

load <vector_name>

When G7 is working on a vector in a Vam file, it pulls the whole time series for the vector into a sort of vector workspace. Usually, this is done implicitly simply by referring to the vector or one of its elements. The load command enables the user to do this loading explicitly. It is used principally in connection with the index command described below.

Example:

load pce

Related Topics: store

look <bank_location>

This command from the G7 console is equivalent to the Bank | Look menu command. As usual, <bank_location> is a letter between ‘a’ and ‘z’ to which you have assigned the bank you would like to look at. For look to work, there must be a stub file created with the same root name as the bank. For example, the stub file for the NIPAA.HBK, databank, which is a hashed bank, is NIPAA.STB.

The look command brings up the stub file in a scrolling list box, from which you can select lines of the file to print out and graph the various series in the bank. If you double-click on the series name, the series will be printed out using the current graph dates (gdates), and the title of the graph will be the line of the stub file. The series also will be printed to the G7 console.

Related Topics: Dates in G7, Stub Files, gdates, graph, tdates

ls [-<flag>] <x> <y> <date1> [<date2>] [direction]

This linkseries command takes series <x> and <y> from the workspace or the assigned bank, calculates their ratio at the specified linking date <date1>. This ratio then is used to move <x> with non-zero entries from series <y>. The linking direction can be ‘f’ for forward or ‘b’ for backward. The default is ‘f’. If the variable <x> is found in the workspace, then the result is stored with name <x> in the workspace. If a bank letter is provided for <x>, and if this bank is the default Vam file, then the modified value of <x> is stored in the Vam file. Otherwise, the result is stored to the workspace. If the optional date <date2> is provided, then linking will extend from the base period <date1> to the terminal period <date2>. <date2> is the last period (if the direction is ‘f’) or first period (if the direction is ‘b’) that will be modified. If not given, then the values will be modified to the end of the series, with no regard for the fdates settings that ordinarily will limit the extent of operations.

If the signs of the values in x and y are not the same, then the result will move in the opposite direction of the guide series; this usually is not desirable and a warning will be given. Three alternatives to the basic routine may be implemented by using a flag.

-f:

the calculation using standard algorithm without reporting problems.

-i:

if x and y are of opposite sign in the base year, then ensure positive correlation between x and y such that

x = a + b*y b = x{date} / y{date} * sign( x{date}*y{date} ) a = x{date} - b * y{date}

-a:

if x and y are of opposite sign in the base year, and the problem is ill-conditioned such that y lies close to zero in the link year, then ensure positive correlation between x and y such that

x = a + b*y b = x{date}/ AVG(y) * sign( x{date}*AVG(y) ) a = x{date} - b * y{date}

where AVG(y) is calculated from fdate1 to <date>, if direction is ‘b’, or from <date> to fdate2, if direction is ‘f’. This may be useful when y{date} is close to zero so that scaling parameter b is large, thus scaling x excessively. If AVG(y) is close to zero, then b = x{date} * sign( x{date}*AVG(y) ). It is the user’s responsibility to ensure that the results are useful.

Related Topics: @bmk function