Summary¶

Inputs¶

Key

Enter the .hex key associated with the data set to be summarized.

X

Select the columns for which summary information is desired. If a header was not used in the original data set, the columns are indexed by number beginning at 0 on the left.

Max Column Display

An alternative to selecting specific X; Max Column Display allows users to define the maximum number of columns that summary information will be provided for. For example, if users specify Max Column Display to be 5, the first 5 columns of the original data set will be displayed, and all others will be omitted regardless of the subset of columns selected in the X field.

Output¶

Column Name

The variable name provided in the header during data parsing, or the default assigned by H2O if no header was included.

Base Stats for Numerical Data

NAs

Displays a count of the number of elements in the column interpreted as NA by H2O, either because the original element was given as NA or because the original element was uninterpretable when the data were parsed.

Average: Abbreviated avg

The arithmetic mean of the data in the column, defined as the sum of the values in the column divided by the number of elements in the column.

\(Average\:(X_{i})=\frac{\sum_{k=1}^{N}X_{k}}{N}\)

Standard Deviation: Abbreviated sd

The standard deviation of the data in the column, defined as the square root of the sum of the deviance of observed values from the mean divided by the number of elements in the column less one. For some columns of data where the standard deviation is a very, very small negative numbe, a standard deviation of -0 may be produced. The exact value has not been rounded, but the number of digits displayed have been truncated.

\(Standard\:deviation\:(X_{i})=\sqrt{\frac{\sum_{k=1}^{N} (X_{k}-\bar X)^2}{N-1}}\)

Zeros:

The number of elements in the column that are 0.

Min[ ] and Max[ ]

Return a list of the minimum and maximum elements of the column respectively. The default is the lowest and highest five numbers, but for binomial data (for example), the minimum are listed as 0,1 and the maximum as 1,0.

Percentiles

A table displayed only for Numeric Data. A percentile of value X at P threshold that is interpreted as “X is greater than or equal to P percent of the values in this column.” For example, if 83 is the value at threshold .75, we would say that 83 is greater than 75% of the other elements in the column.

Threshold

Displays the percentile levels. For example, .25 corresponds to the 25th percentile, .5 corresponds to the median, and .75 corresponds to the 75th percentile.

Value

In H2O, the values presented relative to thresholds as percentiles are obtained through approximation relying on information derived from the histogram.

Base Stats for Categorical Data

NAs

Displays a count of the number of elements in the column interpreted as NA by H2O, either because the original element was given as NA or because the original element was uninterpretable when the data were parsed.

Cardinality

Displays the number of unique categories present in the data. For example, if a column of 100 elements gives information about the color of an object, and the possible values for color are red, green, and blue, cardinality is 3.

Histogram

First Row (top)

Lists the unique values or categorical levels in the column.

Second Row (middle)

Lists a corresponding count of the number of observations of the value given in the top row.

Third Row (bottom)

Gives a percentage of the number of observations that are equal to the corresponding value.

Users should note that the tables in Summary can become quite large. When the width of the table exceeds the width of the display users can view more information by hovering over the table with their mouse and scrolling left and right.