Converts the selected columns content to a JSON value row-wise or column-wise. It also have an option to "undo" -with some limitations- the JSON to Table transformation and create JSON values for each row based on the column names.
Example input table:

a.b	a.c	d
b0	c0	d0
b1	c1	d1

With the different parameters, the following JSON values are generated:
Row-oriented

[ {
  "a.b" : "b0",
  "a.c" : "c0",
  "d" : "d0"
}, {
  "a.b" : "b1",
  "a.c" : "c1",
  "d" : "d1"
} ]

Column-oriented (with Row keys as JSON value with key: "ROWID"):

{
  "ROWID" : [ "Row0", "Row1" ],
  "a.b" : [ "b0", "b1" ],
  "a.c" : [ "c0", "c1" ],
  "d" : [ "d0", "d1" ]
}

Keep rows (with Column names as paths, separator: .):

{
  "a" : {
    "b" : "b0",
    "c" : "c0"
  },
  "d" : "d0"
}

{
  "a" : {
    "b" : "b1",
    "c" : "c1"
  },
  "d" : "d1"
}

Options

Input columns

The selected columns will be transformed.

Row keys|Omit row key

The row keys will be omitted, not used in the generated JSON value (which are arrays when the rows are not kept).

Row keys|Row key as JSON key

The row keys are added to the generated JSON value as a key, in this case not an array, but an object is created.

Row keys|Row key as JSON value with key

The row keys will be included in the generated JSON (array) value with the specified key.

Row-oriented

the values from the selected columns are collected and concatenated row-wise to a JSON object/array, after that these are combined by the rows to a single JSON value.

Column-oriented

the values from the selected columns are collected and concatenated column-wise to a JSON object/array, after that these are combined by the columns to a single JSON value.

Keep rows

the selected columns are combined to a new JSON column containing the values from the columns and the name of the columns as keys. This option does not combine the rows of the input table.

Remove source columns

When checked, the source columns are removed when the rows are kept. It has no effect for the other Aggregation direction options, when all columns are removed.

Column names as paths, where path separator in column names

When From column names as path is selected, this value will be used to find the keys for JSON columns based on the column names.

Missing values are omitted

Missing values from the input table do not generate a key in the resulting JSON structure, they are omitted completely.
Note that in a column-oriented transformation (see above) missing cells will still be inserted a null values in the column's array because otherwise the row arrays for different columns may have different numbers of entries. This would make it impossible to reconstruct the original table

Missing values are inserted as 'null'

Missing values from the input table are insert as null values.

Output column name

Name of the resulting column.

Input Ports

A datatable

Output Ports

Table containing the JSON column

Best Friends (Incoming)

• Column Rename (7 %) Streamable
• Column Filter (7 %) Streamable
• Table Creator (5 %) 
• Row Filter (5 %) Streamable
• Joiner (3 %) 
• Show all 312 recommendations

Best Friends (Outgoing)

• JSON Writer (17 %) 
• POST Request (7 %) Streamable
• JSON to Table (6 %) 
• MongoDB Writer (6 %) 
• Container Output (JSON) (5 %) 
• Show all 196 recommendations

Workflows

• _​01_​TestWF_​Fails_​Then_​Succeeds (KNIME Hub) 
• 00_​call_​local_​workflow_​(JSON) (KNIME Hub) 
• 00_​check_​prime_​number (KNIME Hub) 
• 01_​Bioactivity_​Prediction_​Load_​Local (KNIME Hub) 
• 01_​Bioactivity_​Prediction_​Load_​Pull_​Data (KNIME Hub) 
• Show all 80 workflows

Installation

To use this node in KNIME, install KNIME JSON-Processing from the following update site:

A zipped version of the software site can be downloaded here. Read our FAQs to get instructions about how to install nodes from a zipped update site.

Developers

You want to see the source code for this node? Click the following button and we’ll use our super-powers to find it for you.