Exercises and troubleshooting

You can also download together the three CSVs in a ZIP archive.

The exercises use small synthetic Datasets. Import the supplied CSV files into the Session as tables with enabled numerical columns. Use the first row as the header and the units specified for each exercise, then open Curve Builder.

Interpolate a 2D characteristic curve

Use families.csv. The X/Y coordinates are dimensionless and family_id and member_id are numerical codes. The table has four groups of five lines.

  1. Select Dataset = families, X = x, Y = y and Mode = 2d.
  2. Family variable = family_id, Family value (exact) = 1, Member variable = member_id and Member value (exact) = 0. Activate both Filter enabled boxes. If the same import procedure has produced a different format for the key, select the one shown in the values menu.
  3. Press Add source rows and verify that five rows have been added, with X from 0 to 4 and Y from 1 to 4.2. Press Whole curve and check the five selection dots.
  4. Choose Method = pchip, Target samples = 101 and leave Parameter = variable. With Auto-order off, press Preview. Wait for 101 target points from 5 rows.
  5. Switch to Options → Show selected characteristic / subset and recalculate. Compare legend and count of the rows used. Then enter famiglia_1_membro_0 as a new ID and publish the result.

Fit a time-parameterized 3D path

Use signals.csv. time_s is in seconds, and voltage_v and reference_v are in volts. The file contains 21 rows sampled every half second from 0 to 10 s.

  1. If you’re continuing with the 2D example, press Restart sources. Choose the Dataset signals, Mode = 3d, Parameter = variable, X = time_s, Y = voltage_v, Z = reference_v and t = time_s. Leaves Plot Selection blanks and identity roles, with its filters off.
  2. Press Add source rows, then Whole curve. Set Method = line and Target samples = 101. Press Preview. Check the 21 used rows, the 101 samples and separate statistics with units s, V and V.
  3. Enter a new ID, for example segnali_fitto_3d. Leave Open Plot Dataset overlay after publication selected and publish. Rotate the line in interactive representation and find in the Session the columns t, x, y and z.

In this example the units of X and Y differ. The variable parameter avoids using coordinates with different units in the same length measure.

Review the point order of a planar contour

Use contour.csv, with dimensionless coordinates. The first and last point coincide. The order of the internal points is deliberately irregular.

  1. Start over with Restart sources. Select Dataset contour, X = x, Y = y, Mode = 2d, Parameter = arc_length and Method = linear.
  2. Add rows and use Whole curve. Open Options → Review original / proposed order…. Compare ordinals, alerts and tracks with Compare connectivity…. Check that the last point remains the closing point.
  3. Accept the proposal with Use reviewed order and press Preview. The X or t column now shows cumulative distance. Check both X and Y as calculated coordinates.

Recover from an error

Rows are visible, but Preview ignores them.
Check selected and use Whole curve or Highlighted points.
The plot shows more rows than the count.
Check the plot scope and read the fit rows count.
The plot connects groups that should be separate.
Recheck sources, filter keys and sequence. Use a curve for each physically separate segment.
The Workspace opens, but the plot is missing.
Check the restored sources, any order and refer to Preview.
Session changed appears.
Reopen the Builder or Workspace command in the Active Session. The lines and the Workspace refer to the Session where they were loaded.