Only SELECT is allowed on SWERB. INSERT, UPDATE, and DELETE are not allowed.

Only SELECT is allowed on SWERB. INSERT, UPDATE, and DELETE are not allowed.

Only SELECT is allowed on SWERB_LOCS. INSERT, UPDATE, and DELETE are not allowed.

The format of the data that is inserted into the SWERB_UPLOAD view is complex. This section provides an overview and remarks on unusual features^[284] and the tasks required of the data entry manager to convert the raw SWERB data into an uploadable format. The description of the SWERB_UPLOAD view in the following sections describes how the various uploaded columns map into the columns of Babase's tables. Because, excepting variances described in this section, the uploaded data comes directly from the GPS units used to collect SWERB data the reader should rely on the description of the SWERB data collection protocol in the Amboseli Baboon Research Project Monitoring Guide for a complete description of the data format.^[285]

Each upload into the SWERB_UPLOAD view must consist of the data collected on a single GPS unit by a single observation team during the course of a single day.

The data is uploaded as a collection of lines containing tab-delimited text. Each line represents a waypoint recorded by the operator. The lines are expected to be in chronological order, the first line being the first waypoint recorded and the last line being the last, with the exceptions that the lines contriving any one bout of observation of any one (sub)group must be contiguous and that the begin and end lines which record the sleeping grove must immediately precede the begin and end lines which record the descent or ascent time.^[286] Consequently the following constraints are imposed on the data: the first line(s) must record the observation team's departure from camp; the lines representing a bout of observation must be contiguous; the line denoting the beginning of a bout of observation must precede all of the bout's other lines; the line denoting the end of a bout of observation must follow all of the bout's other lines; the line recording the median descent time, when present, must immediately follow the line denoting beginning of the bout of observation and the previous night's sleeping grove; the line recording the median ascent time, when present, must immediately follow the line denoting end of the bout of observation and the night's sleeping grove; in those cases where a group utilizes more than one sleeping grove the sleeping grove information must consist of contiguous pairs of lines (as just described) with no intervening lines of another sort; notwithstanding anything to the contrary above, lines denoting observations of the non-focal group may appear at any point after the lines representing departure from camp.^[287]

The first 2 lines of the uploaded file are required to be departure lines, lines which record information about the departing observation team. The first line must begin with a D, it lists the observers. The initials supplied on this first line control the SWERB_OBSERVERS.Role value used, the value used being the referenced OBSERVERS.SWERB_Observer_Role column's value. The second line must begin with DD, it lists the drivers. The initials supplied on the second line control the SWERB_OBSERVERS.Role value used, the value used being the referenced OBSERVERS.SWERB_Driver_Role column's value.

It is an error if all the lines representing departure from camp indicate the use of more than one GPS unit. Each data upload into SWERB_DEPARTS must come from a single GPS unit.

It is an error if observer codes cannot be unambiguously interpreted in the departure lines. This means that when there is an observer code which is shorter and match in its entirety the beginning of other observer codes then none of these observer codes, neither the shorter nor the longer, can be reliably used in SWERB departure waypoints.^[288] If there is ambiguity the offending observer code must be manually removed from the departure line and manually inserted into SWERB_OBSERVERS after uploading the data file.

When the field team records coordinates for the start or stop of a bout of observation but somehow fails to record the time, then that time must be estimated by the data entry staff and included in the Description. In this case the columns Timeest and Source must be added to the data file by the data manager. The data manager should supply values for these columns only in the begin and end lines.

When the field team fails to record the start or stop of a bout of observation at all, the data manager needs to create one. In these cases, coordinates cannot be estimated, but the start/stop time may be known or estimated from other data. However, when a date and time are provided (when Description is not NULL), it is normally a rule that coordinates must also be provided (Position cannot be NULL). To manage this conflict, the boolean column BE_Has_Coords is used. When FALSE, the Position must be NULL and the Description is allowed to be non-NULL^[289]. When BE_Has_Coords is TRUE or NULL and the Description is not NULL, the Position cannot be NULL, as usual.

The BE_Has_Coords column is only used for begins and ends of observation bouts. This column must be NULL for all other rows.

The data collection protocols require that each observation team record ascent and descent times and groves for their first and last observations of each group for the day. When more than one observation team observes a single group on a given day then the data manager must choose which observation team's ascent/descent information is to be used.^[290]The unused ascent and descent information must be removed from the uploaded data. This requires removing the grove from the end of the waypoint text, in the Name column, and deleting the line denoting median descent/ascent time. This should leave a single beginning of observation/end of observation line in the place within the file from which the sleeping grove ascent/descent information has been purged.

The SWERB_UPLOAD view treats a leading P character before grove codes written into the uploaded begin and end lines as an indication that the sleeping grove is probable (SWERB_LOC_DATA.Loc_Status is P) unless the result of removing the leading P produces a code which is not in SWERB_GWS as a grove. In this case the leading P is considered part of the grove code.

Although the field operators enter information after the B and E codes in those waypoints recording the beginning and ending of bouts of observation that do not denote descent from or ascent into sleeping groves, the SWERB_UPLOAD view is unable to process this additional information. The data manager must remove this information from those begin and end lines that occur when observation of the focal group is interrupted for some reason.

When the field team records secondary ascents or descents, when there are subgroups and more than one begin or end is recorded for the group, the data managers must add additional lines to the uploaded data to convert these “extra” begins or ends into independent bouts of observation^[291]. All of these lines, the original secondary ascent or descent lines and the additional lines added by the data manager, must be marked with a TRUE value in a Secondary_AD column. (This column will also have to be added by the data manager.)

The bout of observation created for the secondary ascent or descent must consist only of begin and end rows, no other kinds of observation are allowed. The SWERB_UPLOAD view will generate an error if other kinds of observations are interspersed between the begin and end rows of a secondary ascent or descent.^[292]

Secondary ascents and descents must occur during a regular bout of observation -- uploaded rows with a TRUE Secondary_AD value must be preceded by non-secondary begin rows and followed by non-secondary end rows.

Lone animal sightings must be flagged as such in the SWERB_UPLOAD.Lone_Animal column. The sex of the individual must match the sex indicated in the SWERB_UPLOAD.Lone_Animal column.

The SWERB_UPLOAD view looks up the the sname for lone animal entered as part of the garmincode in the Unksname column of the UNKSNAMES table. If found and the related UNKSNAMES.Lonemale value is M the sname of the lone male is stored in SWERB_DATA.Lone_Animal as a NULL.

It is an error to include values in both the Description and the Date columns on the same line.

The geographic coordinates for each row are recorded in the Position column. As discussed elsewhere, the provided coordinates may use WGS 1984 UTM Zone 37South coordinates or longitude and latitude via the WGS 1984 2D CRS, but the location will be stored in its respective table as a WGS 1984 UTM Zone 37South location. When a coordinate in longitude or latitude converts to a UTM coordinate with too many digits after the decimal, the converted value is rounded to the nearest 0.1.

Because the coordinates in the Position column may be from either system, this view uses the provided Position to guess which system is being used. When the provided value begins with 37 M, the coordinates are assumed to be from the WGS 1984 UTM Zone 37South system. The expected format of those data is “37 M” space [X-coordinate] space [Y-coordinate]. The XY units are in meters and are always positive. When the provided value begins with 37., the coordinates are presumed to be from the WGS 1984 2D CRS. The expected format of those data is [longitude] space [latitude]. The units are decimal degrees, with a positive longitude and a negative latitude.

Only INSERT is allowed on SWERB_UPLOAD, SELECT, UPDATE, and DELETE are not allowed. Inserting a row into SWERB_UPLOAD inserts a rows into SWERB tables as described above.