Vector Data Lines |
Top Previous Next |
|
The survey data files you prepare for Walls resemble fieldbook pages in that they consist of lines (or rows) of measurement sets, each defining a vector between two named stations. Since the name pairs are sufficient to determine a survey's connectivity, a successful compilation by Walls does not require that the lines of vector data be ordered in any specific way, or even that vectors be connected (i.e., have matching endpoints). However, if you move lines around via cut-and-paste, you must be careful that each vector has applied to it the correct #Units, #Prefix, and #Segment directives. The following sections describe the components of a vector definition in more detail.
The first two items on compass and tape (CT) and rectangular coordinate (RECT) data lines are the FROM and TO station names, in that order. With CT data, the FROM station is always defined as the station from which the (forward) vertical angle and azimuth readings were actually taken in the field. (Reversing the shot direction after the fact makes it difficult for any program to properly apply instrument corrections or to estimate some kinds of systematic error.) With RECT data, the specified vector is always the displacement of the TO station with respect to the FROM station.
Station names, with their implied or explicit prefixes, must be globally unique within any subset of data files that are to be compiled together. This means that if two names in different files match, they are assumed to refer to the same location. Character case is significant when comparing names, but there is a CASE setting of the #Units directive that can perform case conversions. For information on the use of prefixes to insure station name uniqueness, see #Prefix Directive.
Unprefixed names can have a maximum of eight characters and must not contain any colons, semicolons, commas, pound signs (#), or embedded tabs or spaces. In order to avoid possible problems when printing or when exporting data to other programs, you are encouraged to restrict names in new surveys to numbers with alphabetic prefixes or suffixes (e.g., BR123). Also, you'll probably want to take advantage of the editor's auto-sequencing/prefixing feature, which makes entering such names especially easy.
Immediately following the two station names are the three primary measurements that define a three dimensional vector. For compass and tape data lines, they are the Distance (shot length), magnetic Azimuth (bearing), and Vertical angle (inclination), and they must appear in the order specified by the Order=xxx parameter of the #Units directive, where xxx is some permutation of the letters D, A, and V (DAV is the default). For rectangular (RECT) vector data lines, the three numbers are the true north-relative East, North, and Up displacements, with their default order represented by the #Units setting, Order=ENU. The presumed measurement units are those that have been specified for Distance. (See Measurement Units.)
If these three items are not all present, have the wrong format, or are out of range, Walls will abort compilation and highlight the line in an editor window. There are special cases, however, where the vertical measurement can be omitted entirely to specify a constant zero value. The #Units parameter Order=DA, AD, EN, or NE will allow this. Alternatively, you can enter a string of two or more minus signs (e.g., ---) to make it clear that a measurement was not actually taken. The program will accept this placeholder for missing LRUD measurements and for the nonexistent azimuth readings on pure vertical shots.
To handle a large stash of existing survey data, the SRV format supports optional fourth and fifth measurements on compass and tape (CT) data lines. Walls will interpret these as an instrument height above marked station followed (optionally) by a target height above marked station. If the taping method is instrument-to-target (the default assumption), the effect of these heights is equivalent to adding their difference (IH minus TH) to the computed elevation of the TO station with respect to the FROM station. Note that the #Units directive's INCH vertical adjustment is not necessarily related to these measurements; it is independently applied to all CT data, regardless of whether or not explicit height above station measurements are present.
Earlier versions of this program did not support the four different taping methods. When instrument/target height measurements were encountered, the taping method was naturally assumed to be along the line of sight -- that is, instrument-to-target instead of station-to-station. This has always seemed to me the only logical method to adopt since instrument/target height measurements are necessary only when the line between marked stations is either obstructed or not easily accessible. What finally tilted the balance in favor of support was the requirement to handle underwater surveys, where station-to-station taping is indeed the proper way to represent bearings, taped distances, and depth gauge measurements. Here, depth measurements are treated as if they were IT heights.
For a discussion of the difficulties IT measurements might cause, see Height Measurement Pitfalls.
Occasionally you may want to override the default variance assigned to vectors, whether they be compass and tape (CT), rectangular (RECT), or #FIXed vectors. This variance is used by the adjustment and error analysis portion of Walls and will come into play only when the vector is part of a loop system. The override is in the form of a parenthetical expression following the last measurement (or coordinate) on the data line. For more information, see Variance Assignments.
The SRV format allows the specification of Left-Right-Up-Down passage dimensions -- information that is often found in field books and in data managed by other programs. The optional LRUD specification is a bracketed expression, <L,R,U,D>, following the last measurement on a compass and tape data line -- for example, <10.5,3,5,1>. Note that angle brackets distinguish this from a variance override (parenthetical expression) which can either precede or follow it on the data line. If it simplifies typing, substitute asterisks for the angle brackets. Also, space characters can be used instead of comma delimiters: *10.5 3 5 1*.
To allow for the possibility that the associated vector definition can't provide a realistic facing direction for the LRUD, an azimuth value, can be included as a fifth number in any LRUD expression:
NAM1 NAM2 15 -- -90 <5,10,10,0,325>
A recent addition to the format is an optional trailing "C" (or "c") argument in the LRUD expression -- for example, <1,6,5,1,C>. Currently this flag only effects SVG exports where the option "Draw cross section boxes for C-flagged LRUDs", is specified in the SVG Advanced Settings dialog. For more information see the description of the "w2d Lruds" layer in SVG Layer Definitions.
A missing dimension measurement should be indicated by two adjacent minus signs -- for example, <10,5,--,-->. Unless there are other passage dimension measurements at the station (wall shots or other LRUDs), this may result in zeros assumed for the missing values when passages are drawn. Special values to indicate such things as "passage" at a right-angle bend are not supported. Instead you should consider wall shots (see next section) or use one or more LRUDs with specified facing directions. A given station can have more than one LRUD assignment.
Unfortunately, there doesn't seem to be a generally agreed upon definition for the LRUD numbers written in field books -- for example, whether or not the LRUD is to be associated with the TO station instead of the FROM station (SEF style). Notes are often unclear about the method used, so one has to decide if it's worth tracking down and querying members of the surveying team. Another problem from a software perspective is that we can't reliably convert between the different variants of LRUD that have been proposed. (The SEF format supports only the FROM station variant.) The prospect of recognizing and not just accommodating several incompatible LRUD schemes never did interest me. Eventually, however, it became necessary to support large projects where LRUDs of several kinds had been painstakingly obtained and keyed into computer files.
The #Units directive currently supports four popular LRUD style options: FROM-station-perpendicular (default), TO-station-perpendicular, FROM-station-bisector, and TO-station-bisector. (LRUD=F/T/FB/TB.) You can also specify what order the Left, Right, Up, Down values appear in the expression. For example, LRUD=T:UDLR would specify TO-station LRUDs with the ceiling and floor distances appearing before the left and right distances.
During raw data processing, Walls converts the various LRUD styles to an unambiguous internal format: four distances, a direction (azimuth), and an associated station. If your LRUDs are the TO-station type, it's very important that you use the LRUD=T (or TB) units specification, perhaps as a Compile Options entry in the project root folder's Compile Options dialog. Otherwise, LRUDs will be associated with the wrong station.
You'll find, however, that the wall rendering algorithm is fairly robust and can be easily adjusted after the fact to prevent passage narrowing at turns. Although Walls-specific, there is also the option to include an overriding azimuth in the LRUD expression as described above.
With either of the two basic LRUD methods, TO-station or FROM-station, you'll sometimes need to specify starting station LRUDs or ending station LRUDs. Depending on the method, you'll need sequences like the following:
#Units LRUD=T A0 <2.5,5,3,0> ;Starting station LRUD A0 A1 30 275 -5 <0,7,4,0> --- etc.
#Units LRUD=F --- A0 A1 30 275 -5 <0,7,4,0> A1 <2.5,5,3,0> ;Ending station LRUD --- etc.
Finally, at any point in your data files you can associate a station with LRUD dimensions and a facing azimuth:
A1 <2,3,5,0,275>
LRUDs are included as shapefile attributes and optionally as part of exported SVG files. They can also be displayed on printed and screen maps. For more information, see Passage Display Options.
Another method of specifying passage dimension data is available which is more flexible and less ambiguous than conventional LRUDs. These are simply shots to anonymous points on a passage boundary:
A1 - 10 210 A1 - 12 300 +10 - A1 0.5 45 A1 - 3.5 -- +90 --- etc.
The main distinction of a wall shot is that a minus sign (- or --) is used for one of the names. The inclination is optional, and the processed data will be used only for the representation of walls. Like LRUDs, they don't contribute to survey statistics nor do they appear on the preview map. The program will throw them into the mix with LRUDs when defining passage outlines for plans and profiles. Reasonably accurate outlines of large rooms and passage junctions can be achieved this way when vector-associated LRUDs provide too little information.
There is only one rule to follow with wall shots. For an accurate plan or profile view, there should be at least one wall shot (or LRUD) with the unnamed station residing at a point farthest in horizontal or vertical extent in the specified direction. There's no harm, however, in shooting to arbitrary points on the floor or ceiling. The Walls3D viewer has a relatively new option to display wall shots and LRUDs.
Segment Assignment for a Specific Vector Finally, an optional segment specification, prefixed by the token "#S" (or "#Seg, or #Segment), can appear as the last data item of any line that defines a vector -- namely #FIXed stations, RECT coordinates, and compass and tape data lines. It serves to assign the segment attribute of a single vector without changing the current default segment. Currently, this is the only directive that can appear on the same line as a vector's data. A common example is a flag-like segment assignment, such as "#S L", to indicate that the vector should be excluded from length calculations. Such vectors could then be "detached" from the segment tree prior to statistics being calculated. (See Segments Page.)
|