DataFrame.loc

Access a group of rows and columns by label(s)

danfo.DataFrame.loc(args)

Parameters
Type
Description
Default

args

Object

{

rows: Array, labels, Boolean mask of row index

columns: Array, labels of column names

}

Examples

.loc() is label position based-from 0 to length-1 of the row axis.

Allowed inputs for are:

  • An integer, e.g. "r1".

  • A list or array of integers, e.g. ["a", "b", "d"].

  • A boolean mask. E.g [ true, false, false ]

  • A string slice object with ints, e.g. ['"a":"d"'], ["1:4"]

Note: only **** the start label is included, and the end label is ignored.

.loc will raise a ValueEror if a requested label is not found.

Index by specific rows and return all columns

If the row's index is specified and the columns are not, then it returns all columns and just the specified rows.

const dfd = require("danfojs-node")

let data = { "Name": ["Apples", "Mango", "Banana", "Pear"],
            "Count": [21, 5, 30, 10],
            "Price": [200, 300, 40, 250] }

let df = new dfd.DataFrame(data, {index: ["a", "b", "c", "d"]})
df.print()
let sub_df = df.loc({rows: ["a", "c"]})
sub_df.print()

Index by a list of column names and return all rows

Index both axes by the specified labels

Index by a slice of row

The loc function also accepts string slices of the form [start: end], e.g [`"a":"c"`]. This will return all values from label positions a to c.

Note that when using loc with alphabetic slices. We expect you to pass labels in the correct format. That is, string labels must be explicitly quoted. For example, the following loc slice will throw an error: df.loc({ row: ["a:e"]}).print() For the slice above to work, you must quote each slice, e.g: df.loc({ row: [`"a":"e"`]}).print() Inner quotes are not needed for numeric indices!

Slice DataFrame rows by boolean condition

Slice DataFrame rows by multiple boolean conditions

By design, you can chain as many boolean logic as required, as long as they resolve to a Boolean array of the same length as the DataFrame.

Slice DataFrame with boolean mask

You can index a DataFrame with an array of boolean values as long as they resolve to an array of the same length as the DataFrame.

PreviousDataFrame.indexNextDataFrame.iloc

Last updated

Was this helpful?