DataFrame

Construct an empty data frame.

 > DataFrame<Object> df = new DataFrame<>();
 > df.isEmpty();
 true 

Construct an empty data frame with the specified columns.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.columns();
 [name, value] 

Construct an empty data frame with the specified columns.

 > List<String> columns = new ArrayList<>();
 > columns.add("name");
 > columns.add("value");
 > DataFrame<Object> df = new DataFrame<>(columns);
 > df.columns();
 [name, value] 

Construct a data frame containing the specified rows and columns.

 > List<String> rows = Arrays.asList("row1", "row2", "row3");
 > List<String> columns = Arrays.asList("col1", "col2");
 > DataFrame<Object> df = new DataFrame<>(rows, columns);
 > df.get("row1", "col1");
 null 

Construct a data frame from the specified list of columns.

 > List<List<Object>> data = Arrays.asList(
 >       Arrays.<Object>asList("alpha", "bravo", "charlie"),
 >       Arrays.<Object>asList(1, 2, 3)
 > );
 > DataFrame<Object> df = new DataFrame<>(data);
 > df.row(0);
 [alpha, 1] 

Construct a new data frame using the specified data and indices.

Add a new column to the data frame containing the value provided. Any existing rows with indices greater than the size of the specified column data will have null values for the new column.

 > DataFrame<Object> df = new DataFrame<>();
 > df.add("value", Arrays.<Object>asList(1));
 > df.columns();
 [value] 

Add the results of applying a row-wise function to the data frame as a new column.

Add the list of values as a new column.

Add new columns to the data frame. Any existing rows will have null values for the new columns.

 > DataFrame<Object> df = new DataFrame<>();
 > df.add("value");
 > df.columns();
 [value] 

Apply an aggregate function to each group or the entire data frame if the data is not grouped.

Append rows indexed by the the specified name to the data frame.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.append("row1", Arrays.asList("alpha", 1));
 > df.append("row2", Arrays.asList("bravo", 2));
 > df.index();
 [row1, row2] 

Append rows to the data frame.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.append(Arrays.asList("alpha", 1));
 > df.append(Arrays.asList("bravo", 2));
 > df.length();
 2 

Apply a function to each value in the data frame.

 > DataFrame<Number> df = new DataFrame<>(
 >         Arrays.<List<Number>>asList(
 >                 Arrays.<Number>asList(1, 2),
 >                 Arrays.<Number>asList(3, 4)
 >             )
 >     );
 > df = df.apply(new Function<Number, Number>() {
 >         public Number apply(Number value) {
 >             return value.intValue() * value.intValue();
 >         
 >     });
 > df.flatten();
 [1, 4, 9, 16] }

Cast this data frame to the specified type.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.append(Arrays.asList("one", "1"));
 > DataFrame<String> dfs = df.cast(String.class);
 > dfs.get(0, 0).getClass().getName();
 java.lang.String 

Update the data frame in place by overwriting any null values with any non-null values provided by the data frame arguments.

Return a data frame column as a list.

 > DataFrame<Object> df = new DataFrame<>(
 >         Collections.emptyList(),
 >         Arrays.asList("name", "value"),
 >         Arrays.asList(
 >             Arrays.<Object>asList("alpha", "bravo", "charlie"),
 >             Arrays.<Object>asList(1, 2, 3)
 >         )
 >     );
 > df.col("value");
 [1, 2, 3] 

Return a data frame column as a list.

 > DataFrame<Object> df = new DataFrame<>(
 >         Collections.emptyList(),
 >         Arrays.asList("name", "value"),
 >         Arrays.asList(
 >             Arrays.<Object>asList("alpha", "bravo", "charlie"),
 >             Arrays.<Object>asList(1, 2, 3)
 >         )
 >     );
 > df.col(1);
 [1, 2, 3] 

Return the column names for the data frame.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.columns();
 [name, value] 

Concatenate the specified data frames with this data frame and return the result.

 > DataFrame<Object> left = new DataFrame<>("a", "b", "c");
 > left.append("one", Arrays.asList(1, 2, 3));
 > left.append("two", Arrays.asList(4, 5, 6));
 > left.append("three", Arrays.asList(7, 8, 9));
 > DataFrame<Object> right = new DataFrame<>("a", "b", "d");
 > right.append("one", Arrays.asList(10, 20, 30));
 > right.append("two", Arrays.asList(40, 50, 60));
 > right.append("four", Arrays.asList(70, 80, 90));
 > left.concat(right).length();
 6 

Convert columns based on the requested types.

Note, the conversion process replaces existing values with values of the converted type.

 > DataFrame<Object> df = new DataFrame<>("a", "b", "c");
 > df.append(Arrays.asList("one", 1, 1.0));
 > df.append(Arrays.asList("two", 2, 2.0));
 > df.convert(
 >     null,         // leave column "a" as is
 >     Long.class,   // convert column "b" to Long
 >     Number.class  // convert column "c" to Double
 > );
 > df.types();
 [class java.lang.String, class java.lang.Long, class java.lang.Double] 

Attempt to infer better types for object columns.

The following conversions are performed where applicable:

• Floating point numbers are converted to Double values
• Whole numbers are converted to Long values
• True, false, yes, and no are converted to Boolean values
• Date strings in the following formats are converted to Date values:
  2000-01-01T00:00:00+1, 2000-01-01T00:00:00EST, 2000-01-01
• Time strings in the following formats are converted to Date values:
  2000/01/01, 1/01/2000, 12:01:01 AM, 23:01:01, 12:01 AM, 23:01

Note, the conversion process replaces existing values with values of the converted type.

 > DataFrame<Object> df = new DataFrame<>("name", "value", "date");
 > df.append(Arrays.asList("one", "1", new Date()));
 > df.convert();
 > df.types();
 [class java.lang.String, class java.lang.Long, class java.util.Date] 

Draw the numeric columns of this data frame as a chart in the specified Container using the specified type.

Draw the numeric columns of this data frame as a chart in the specified Container.

Create a new data frame by leaving out the specified columns.

 > DataFrame<Object> df = new DataFrame<>("name", "value", "category");
 > df.drop(2).columns();
 [name, value] 

Create a new data frame by leaving out the specified columns.

 > DataFrame<Object> df = new DataFrame<>("name", "value", "category");
 > df.drop("category").columns();
 [name, value] 

Return a map of group names to data frame for grouped data frames. Observe that for this method to have any effect a groupBy call must have been done before.

Returns a view of the of data frame with NA's replaced with fill.

Return the values of the data frame as a flat list.

 > DataFrame<String> df = new DataFrame<>(
 >         Arrays.asList(
 >                 Arrays.asList("one", "two"),
 >                 Arrays.asList("alpha", "bravo")
 >             )
 >     );
 > df.flatten();
 [one, two, alpha, bravo] 

Return the value located by the (row, column) coordinates.

 > DataFrame<Object> df = new DataFrame<Object>(
 >     Collections.emptyList(),
 >     Arrays.asList("name", "value"),
 >     Arrays.asList(
 >         Arrays.asList("alpha", "bravo", "charlie"),
 >         Arrays.asList(10, 20, 30)
 >     )
 > );
 > df.get(1, 0);
 bravo 

Return the value located by the (row, column) names.

 > DataFrame<Object> df = new DataFrame<Object>(
 >     Arrays.asList("row1", "row2", "row3"),
 >     Arrays.asList("name", "value"),
 >     Arrays.asList(
 >         Arrays.asList("alpha", "bravo", "charlie"),
 >         Arrays.asList(10, 20, 30)
 >     )
 > );
 > df.get("row2", "name");
 bravo 

Group the data frame rows by the specified column names.

Group the data frame rows by the specified columns.

Group the data frame rows using the specified key function.

Return a data frame containing the first limit rows of this data frame.

 > DataFrame<Integer> df = new DataFrame<>("value");
 > for (int i = 0; i < 20; i++)
 >     df.append(Arrays.asList(i));
 > df.head(3)
 >   .col("value");
 [0, 1, 2] 

Return a data frame containing the first ten rows of this data frame.

 > DataFrame<Integer> df = new DataFrame<>("value");
 > for (int i = 0; i < 20; i++)
 >     df.append(Arrays.asList(i));
 > df.head()
 >   .col("value");
 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9] 

Return the index names for the data frame.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.append("row1", Arrays.asList("one", 1));
 > df.index();
 [row1] 

Return true if the data frame contains no data.

 > DataFrame<Object> df = new DataFrame<>();
 > df.isEmpty();
 true 

Create a new data frame containing boolean values such that null object references in the original data frame yield true and valid references yield false.

 > DataFrame<Object> df = new DataFrame<Object>(
 >     Arrays.asList(
 >         Arrays.asList("alpha", "bravo", null),
 >         Arrays.asList(null, 2, 3)
 >     )
 > );
 > df.isnull().row(0);
 [false, true] 

Return an iterator over the rows of the data frame. Also used implicitly with foreach loops.

 > DataFrame<Integer> df = new DataFrame<>(
 >         Arrays.asList(
 >             Arrays.asList(1, 2),
 >             Arrays.asList(3, 4)
 >         )
 >     );
 > List<Integer> results = new ArrayList<>();
 > for (List<Integer> row : df)
 >     results.add(row.get(0));
 > results;
 [1, 2] 

Return a new data frame created by performing a left outer join of this data frame with the argument and using the row indices as the join key.

 > DataFrame<Object> left = new DataFrame<>("a", "b");
 > left.append("one", Arrays.asList(1, 2));
 > left.append("two", Arrays.asList(3, 4));
 > left.append("three", Arrays.asList(5, 6));
 > DataFrame<Object> right = new DataFrame<>("c", "d");
 > right.append("one", Arrays.asList(10, 20));
 > right.append("two", Arrays.asList(30, 40));
 > right.append("four", Arrays.asList(50, 60));
 > left.join(right)
 >     .index();
 [one, two, three] 

Return a new data frame created by performing a left outer join of this data frame with the argument using the specified key function.

Return a new data frame created by performing a join of this data frame with the argument using the specified join type and using the row indices as the join key.

Return a new data frame created by performing a join of this data frame with the argument using the specified join type and the specified key function.

Return a new data frame created by performing a join of this data frame with the argument using the specified join type and the column values as the join key.

Return a new data frame created by performing a left outer join of this data frame with the argument using the column values as the join key.

Return a new data frame created by performing a left outer join of this data frame with the argument using the column values as the join key.

Return a new data frame created by performing a join of this data frame with the argument using the specified join type and the column values as the join key.

Return the length (number of rows) of the data frame.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.append(Arrays.asList("alpha", 1));
 > df.append(Arrays.asList("bravo", 2));
 > df.append(Arrays.asList("charlie", 3));
 > df.length();
 3 

Entry point to joinery as a command line tool. The available commands are:

show

display the specified data frame as a swing table

plot

display the specified data frame as a chart

compare

merge the specified data frames and output the result

shell

launch an interactive javascript shell for exploring data

Return a map of index names to rows.

 > DataFrame<Integer> df = new DataFrame<>("value");
 > df.append("alpha", Arrays.asList(1));
 > df.append("bravo", Arrays.asList(2));
 > df.map();
 {alpha=[1], bravo=[2]}

Compute the mean of the numeric columns for each group or the entire data frame if the data is not grouped.

 > DataFrame<Integer> df = new DataFrame<>("value");
 > df.append("one", Arrays.asList(1));
 > df.append("two", Arrays.asList(5));
 > df.append("three", Arrays.asList(3));
 > df.append("four",  Arrays.asList(7));
 > df.mean().col(0);
 [4.0] 

Return a new data frame created by performing a join of this data frame with the argument using the specified join type and the common, non-numeric columns from each data frame as the join key.

Return a new data frame created by performing a left outer join of this data frame with the argument using the common, non-numeric columns from each data frame as the join key.

Return a data frame containing only columns with non-numeric data.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.append(Arrays.asList("one", 1));
 > df.append(Arrays.asList("two", 2));
 > df.nonnumeric().columns();
 [name] 

Create a new data frame containing boolean values such that valid object references in the original data frame yield true and null references yield false.

 > DataFrame<Object> df = new DataFrame<>(
 >     Arrays.asList(
 >         Arrays.<Object>asList("alpha", "bravo", null),
 >         Arrays.<Object>asList(null, 2, 3)
 >     )
 > );
 > df.notnull().row(0);
 [true, false] 

Return a data frame containing only columns with numeric data.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.append(Arrays.asList("one", 1));
 > df.append(Arrays.asList("two", 2));
 > df.numeric().columns();
 [value] 

Compute the percentile of the numeric columns for each group or the entire data frame if the data is not grouped.

 > DataFrame<Integer> df = new DataFrame<>("value");
 > df.append("one", Arrays.asList(1));
 > df.append("two", Arrays.asList(5));
 > df.append("three", Arrays.asList(3));
 > df.append("four",  Arrays.asList(7));
 > df.mean().col(0);
 [4.0] 

Display the numeric columns of this data frame as a chart in a new swing frame using the specified type.

 > DataFrame<Object> df = new DataFrame<Object>(
 >     Collections.emptyList(),
 >     Arrays.asList("name", "value"),
 >     Arrays.asList(
 >         Arrays.asList("alpha", "bravo", "charlie"),
 >         Arrays.asList(10, 20, 30)
 >     )
 > );
 > df.plot(PlotType.AREA);
  

Display the numeric columns of this data frame as a line chart in a new swing frame.

 > DataFrame<Object> df = new DataFrame<Object>(
 >     Collections.emptyList(),
 >     Arrays.asList("name", "value"),
 >     Arrays.asList(
 >         Arrays.asList("alpha", "bravo", "charlie"),
 >         Arrays.asList(10, 20, 30)
 >     )
 > );
 > df.plot();
  

Compute the product of the numeric columns for each group or the entire data frame if the data is not grouped.

 > DataFrame<Object> df = new DataFrame<>(
 >         Collections.emptyList(),
 >         Arrays.asList("name", "value"),
 >         Arrays.asList(
 >                 Arrays.<Object>asList("alpha", "alpha", "alpha", "bravo", "bravo"),
 >                 Arrays.<Object>asList(1, 2, 3, 4, 5)
 >             )
 >     );
 > df.groupBy("name")
 >   .prod()
 >   .col("value");
 [6.0, 20.0] 

Read the specified csv file and return the data as a data frame.

Read csv records from an input stream and return the data as a data frame.

Read data from the provided query results into a new data frame.

Execute the SQL query and return the results as a new data frame.

 > Connection c = DriverManager.getConnection("jdbc:derby:memory:testdb;create=true");
 > c.createStatement().executeUpdate("create table data (a varchar(8), b int)");
 > c.createStatement().executeUpdate("insert into data values ('test', 1)");
 > DataFrame.readSql(c, "select * from data").flatten();
 [test, 1] 

Read data from the specified excel workbook into a new data frame.

Read data from the input stream as an excel workbook into a new data frame.

Re-index the rows of the data frame using the specified column index, optionally dropping the column from the data.

 > DataFrame<Object> df = new DataFrame<>("one", "two");
 > df.append("a", Arrays.asList("alpha", 1));
 > df.append("b", Arrays.asList("bravo", 2));
 > df.reindex(0, true)
 >   .index();
 [alpha, bravo] 

Re-index the rows of the data frame using the specified column indices and dropping the columns from the data.

 > DataFrame<Object> df = new DataFrame<>("one", "two");
 > df.append("a", Arrays.asList("alpha", 1));
 > df.append("b", Arrays.asList("bravo", 2));
 > df.reindex(0)
 >   .index();
 [alpha, bravo] 

Re-index the rows of the data frame using the specified column indices, optionally dropping the columns from the data.

 > DataFrame<Object> df = new DataFrame<>("one", "two", "three");
 > df.append("a", Arrays.asList("alpha", 1, 10));
 > df.append("b", Arrays.asList("bravo", 2, 20));
 > df.reindex(new Integer[] { 0, 1 , true)
 >   .index();
 [[alpha, 1], [bravo, 2]] }

Re-index the rows of the data frame using the specified column names and removing the columns from the data.

 > DataFrame<Object> df = new DataFrame<>("one", "two");
 > df.append("a", Arrays.asList("alpha", 1));
 > df.append("b", Arrays.asList("bravo", 2));
 > df.reindex("one", true)
 >   .index();
 [alpha, bravo] 

Re-index the rows of the data frame using the specified column names, optionally dropping the columns from the data.

 > DataFrame<Object> df = new DataFrame<>("one", "two", "three");
 > df.append("a", Arrays.asList("alpha", 1, 10));
 > df.append("b", Arrays.asList("bravo", 2, 20));
 > df.reindex(new String[] { "one", "two" , true)
 >   .index();
 [[alpha, 1], [bravo, 2]] }

Re-index the rows of the data frame using the specified column name, optionally dropping the row from the data.

 > DataFrame<Object> df = new DataFrame<>("one", "two");
 > df.append("a", Arrays.asList("alpha", 1));
 > df.append("b", Arrays.asList("bravo", 2));
 > df.reindex("one", true)
 >   .index();
 [alpha, bravo] 

Return a new data frame with the default index, rows names will be reset to the string value of their integer index.

 > DataFrame<Object> df = new DataFrame<>("one", "two");
 > df.append("a", Arrays.asList("alpha", 1));
 > df.append("b", Arrays.asList("bravo", 2));
 > df.resetIndex()
 >   .index();
 [0, 1] 

Reshape a data frame to the specified dimensions.

 > DataFrame<Object> df = new DataFrame<>("0", "1", "2");
 > df.append("0", Arrays.asList(10, 20, 30));
 > df.append("1", Arrays.asList(40, 50, 60));
 > df.reshape(3, 2)
 >   .length();
 3 

Reshape a data frame to the specified indices.

 > DataFrame<Object> df = new DataFrame<>("0", "1", "2");
 > df.append("0", Arrays.asList(10, 20, 30));
 > df.append("1", Arrays.asList(40, 50, 60));
 > df.reshape(Arrays.asList("0", "1", "2"), Arrays.asList("0", "1"))
 >   .length();
 3 

Create a new data frame containing only the specified columns.

 DataFrame<Object> df = new DataFrame<>("name", "value", "category");
  df.retain(0, 2).columns();
 [name, category] 

Create a new data frame containing only the specified columns.

 > DataFrame<Object> df = new DataFrame<>("name", "value", "category");
 > df.retain("name", "category").columns();
 [name, category] 

Return a data frame row as a list.

 > DataFrame<Object> df = new DataFrame<>(
 >         Arrays.asList("row1", "row2", "row3"),
 >         Collections.emptyList(),
 >         Arrays.asList(
 >             Arrays.<Object>asList("alpha", "bravo", "charlie"),
 >             Arrays.<Object>asList(1, 2, 3)
 >         )
 >     );
 > df.row("row2");
 [bravo, 2] 

Return a data frame row as a list.

 > DataFrame<Object> df = new DataFrame<>(
 >         Collections.emptyList(),
 >         Collections.emptyList(),
 >         Arrays.asList(
 >             Arrays.<Object>asList("alpha", "bravo", "charlie"),
 >             Arrays.<Object>asList(1, 2, 3)
 >         )
 >     );
 > df.row(1);
 [bravo, 2] 

Select a subset of the data frame using a predicate function.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > for (int i = 0; i < 10; i++)
 >     df.append(Arrays.asList("name" + i, i));
 > df.select(new Predicate<Object>() {
 >         @Override
 >         public Boolean apply(List<Object> values) {
 >             return Integer.class.cast(values.get(1)).intValue() % 2 == 0;
 >         
 >     })
 >   .col(1);
 [0, 2, 4, 6, 8] } 

Set the value located by the names (row, column).

 > DataFrame<Object> df = new DataFrame<>(
 >        Arrays.asList("row1", "row2"),
 >        Arrays.asList("col1", "col2")
 >     );
 > df.set("row1", "col2", new Integer(7));
 > df.col(1);
 [7, null] 

Set the value located by the coordinates (row, column).

 > DataFrame<Object> df = new DataFrame<>(
 >        Arrays.asList("row1", "row2"),
 >        Arrays.asList("col1", "col2")
 >     );
 > df.set(1, 0, new Integer(7));
 > df.col(0);
 [null, 7] 

Return the size (number of columns) of the data frame.

 > DataFrame<Object> df = new DataFrame<>("name", "value");
 > df.size();
 2 

Compute the standard deviation of the numeric columns for each group or the entire data frame if the data is not grouped.

 > DataFrame<Object> df = new DataFrame<>(
 >         Collections.emptyList(),
 >         Arrays.asList("name", "value"),
 >         Arrays.asList(
 >                 Arrays.<Object>asList("alpha", "alpha", "alpha", "bravo", "bravo", "bravo"),
 >                 Arrays.<Object>asList(1, 2, 3, 4, 6, 8)
 >             )
 >     );
 > df.groupBy("name")
 >   .stddev()
 >   .col("value");
 [1.0, 2.0] 

Compute the sum of the numeric columns for each group or the entire data frame if the data is not grouped.

 > DataFrame<Object> df = new DataFrame<>(
 >         Collections.emptyList(),
 >         Arrays.asList("name", "value"),
 >         Arrays.asList(
 >                 Arrays.<Object>asList("alpha", "alpha", "alpha", "bravo", "bravo"),
 >                 Arrays.<Object>asList(1, 2, 3, 4, 5)
 >             )
 >     );
 > df.groupBy("name")
 >   .sum()
 >   .col("value");
 [6.0, 9.0] 

Return a data frame containing the last ten rows of this data frame.

 > DataFrame<Integer> df = new DataFrame<>("value");
 > for (int i = 0; i < 20; i++)
 >     df.append(Arrays.asList(i));
 > df.tail()
 >   .col("value");
 [10, 11, 12, 13, 14, 15, 16, 17, 18, 19] 

Return a data frame containing the last limit rows of this data frame.

 > DataFrame<Integer> df = new DataFrame<>("value");
 > for (int i = 0; i < 20; i++)
 >     df.append(Arrays.asList(i));
 > df.tail(3)
 >   .col("value");
 [17, 18, 19] 

Copy the values of contained in the data frame into a flat array of length #size() * #length().

Copy the values of contained in the data frame into a array of the specified type. If the type specified is a two dimensional array, for example double[][].class, a row-wise copy will be made.

Copy the values of contained in the data frame into the specified array. If the length of the provided array is less than length #size() * #length() a new array will be created.

Encodes the DataFrame as a model matrix, converting nominal values to dummy variables but does not add an intercept column. More methods with additional parameters to control the conversion to the model matrix are available in the Conversion class.

Encodes the DataFrame as a model matrix, converting nominal values to dummy variables but does not add an intercept column. More methods with additional parameters to control the conversion to the model matrix are available in the Conversion class.

Transpose the rows and columns of the data frame.

 > DataFrame<String> df = new DataFrame<>(
 >         Arrays.asList(
 >                 Arrays.asList("one", "two"),
 >                 Arrays.asList("alpha", "bravo")
 >             )
 >     );
 > df.transpose().flatten();
 [one, alpha, two, bravo] 

Return the types for each of the data frame columns.

Update the data frame in place by overwriting the any values with the non-null values provided by the data frame arguments.

Write the data from this data frame to the specified file as comma separated values.

Write the data from this data frame to the provided output stream as comma separated values.

Write the data from the data frame to a database by executing the provided prepared SQL statement.

Write the data from the data frame to a database by executing the specified SQL statement.

Write the data from the data frame to the provided output stream as an excel workbook.

Write the data from the data frame to the specified file as an excel workbook.