In this lesson

Most of the time, opening a file is a simple drag and drop. This lesson also covers the cases where you need more: many files, or a table whose columns tsmap has to be told about.

• Open files from inside the app, and build one lot from several files
• Choose and rename tests in the test selector
• Map the columns of a CSV, JSON or Parquet file

The quickest start

Load sample data opens an example lot: 25 wafers from five process corners. More sample lots, one per feature, are listed in the user guide.

What tsmap reads

Whatever test data you have, you can probably open it in tsmap.

• STDF and ATDF: tester files, read directly
• CSV, JSON and Parquet: tables, after a simple column mapping step
• .gz, .bz2 and .zip: compressed files are read without unpacking first

Drag and drop

Drop one or more files anywhere in the window. It is the same as picking them. In the desktop app, dropping a folder scans it.

Open files ▾

Open files ▾ opens a menu that offers Choose files… or Scan a folder…

Recent files

Desktop app only.

The last 8 file sets you opened are listed on the start screen, with when you opened each. Click one to open it again, or × to forget it.

Select, then load

The usual routine: search for a string like EDGE, Select all of what is shown, then Load selected…. You load only the files you want.

Filters you use again

• Save filter… keeps the filters in a file; Load filter… reselects matching rows
• A mixed folder: buttons above the table pick which kind of file to load
• Pick more than five files with Open files and tsmap offers this filter table

One lot from separate files

The two files of EDGE-LOT-01 hold one wafer each, and they load as one lot, EDGE-LOT-01. Files that name the same lot open straight away, as long as no wafer appears in more than one of them. A lot tested in several sessions arrives like this, one file per session. The six files used here are in the scan-folder zip on the sample files page.

Add files

Add files appends wafers to what is already loaded. The confirmation warns when the new wafers differ from the ones you have: another die count, other bins, or a wafer ID that's already loaded. Read it and confirm before you continue, because mixing unlike wafers can mislead.

The lot grows

Once the third file is added, its wafer joins the same gallery, Summary and yield.

Naming the wafers

When the files come from different lots, tsmap suggests a label per wafer. A plain ID such as W01 gets its lot in front, so the W01 of EDGE-LOT-01 and the W01 of HY-LOT-04 stay separate, instead of being mistaken for one wafer. Edit any label, then Continue →

Test units across files

When files are combined, a test can arrive in different units.

• One file in mV and another in V? tsmap converts the later file to the first file's unit, and the log says so
• Any other difference, mV against MV included, may be another measurement. That test is left out of the views that compare wafers, so unlike numbers are never compared.

Two more ways to open test data

• A link that tells the web app which files to load
• The command line, which opens the files you name in the desktop app

Both are in lesson 12.

The test selector

A test program can have hundreds or thousands of tests, and you rarely need them all at once. The test selector is where you choose what you need to analyse. A small lot imports every test by default, and Setup ▾ → Tests… opens the selector at any time.

Edit limits…

Edit limits… in the selector opens a table of every test, where you can change its name, limits, units and type before the tests are imported. Lesson 02 explains test and spec limits, then walks through the table.

Big lots start with no tests selected

Every test you import costs memory, so a large lot asks you to choose before it loads.

• A lot with many tests opens the selector first, nothing ticked
• The footer estimates memory: amber is slow, red asks first

Many files, many tests

• Several files: Scan all N files lists tests from every file
• Later, removing a test applies at once; adding one rereads the files

Rename a test

Test names from a tester can be cryptic. Click into a name and type a clearer one. Enter keeps it; only the displayed name changes, not the data.

Save and load definitions

4 Save definitions writes the selection, names, limits, units and test type to a CSV. 5 Load definitions reapplies it, which also gives limits to a program that shipped without them.

Not for STDF or ATDF

STDF and ATDF name every field in their records, so they go straight to the test selector without any need for column mapping.

A CSV, JSON or Parquet table can lay its columns out any way, so tsmap has to ask which column is the X position, which is the bin, and so on.

Column mapping

Every column gets a role. Common names are detected for you, so most of the time you only need to check that they are right, then Continue →

Wide format

In wide format there is one column per test. Each test column takes the Test value role 1.

Long format

In long format there is one row per die per test: a test number or name column, and a result column.

Roles worth knowing

Most columns are a test value or something to ignore. A few roles do more.

• Lot ID: map it so two lots' W01 stay two separate wafers
• Pass/fail test: a PASS/FAIL column, chosen for you
• Display info: a wafer fact, such as temperature, to group by
• X and Y are optional: without them, dies have no position

Many columns at once

1 Filter the columns by name, then give every shown column one role: 2 Test value or 3 Ignore.

Pass bins

Pass bin(s) says which hard bins count as a pass, and the yield figure follows it. The default is 1. If your program passes in another bin, or several, list them with commas, such as 1,7.

Saved mappings

The roles you confirm are remembered for that set of columns, so the next file with the same columns needs no checking. Reset to auto-detected forgets them.

Bin names

A file may carry only bin numbers. Tables show them as bare numbers until you name them in Bin definitions… or load a bin definitions file, which also names the bins in the legend. Lesson 04 covers the window and lesson 12 the files.

The whole app follows

Maps, charts and panels all take the theme's colours.

The log panel

The log panel shows what each load did. Warnings are counted on the Log button until you open it.

Copy the log

Right-click in the log for Select all and Copy all, which puts every entry on the clipboard: the thing to paste into a bug report.

Recap

• Drop files, Open files, or scan a folder and filter by lot
• Add files grows the lot; Setup ▾ → Tests… chooses and renames tests
• Tables need a column mapping, and tsmap remembers it

With your data loaded and your tests chosen, you are ready to read the maps.

More in the user guide: Loading your data