StatsLibpublic final class StatsLib extends Object
| Constructors Summary |
|---|
private StatsLib()
|
| Methods Summary |
|---|
| public static double | avedev(double[] v)returns the mean of deviations from mean.
double r = 0;
double m = 0;
double s = 0;
for (int i=0, iSize=v.length; i<iSize; i++) {
s += v[i];
}
m = s / v.length;
s = 0;
for (int i=0, iSize=v.length; i<iSize; i++) {
s += Math.abs(v[i]-m);
}
r = s / v.length;
return r;
| | public static double | devsq(double[] v)
double r = Double.NaN;
if (v!=null && v.length >= 1) {
double m = 0;
double s = 0;
int n = v.length;
for (int i=0; i<n; i++) {
s += v[i];
}
m = s / n;
s = 0;
for (int i=0; i<n; i++) {
s += (v[i]- m) * (v[i] - m);
}
r = (n == 1)
? 0
: s;
}
return r;
| | public static double | kthLargest(double[] v, int k)returns the kth largest element in the array. Duplicates
are considered as distinct values. Hence, eg.
for array {1,2,4,3,3} & k=2, returned value is 3.
k <= 0 & k >= v.length and null or empty arrays
will result in return value Double.NaN
double r = Double.NaN;
k--; // since arrays are 0-based
if (v!=null && v.length > k && k >= 0) {
Arrays.sort(v);
r = v[v.length-k-1];
}
return r;
| | public static double | kthSmallest(double[] v, int k)returns the kth smallest element in the array. Duplicates
are considered as distinct values. Hence, eg.
for array {1,1,2,4,3,3} & k=2, returned value is 1.
k <= 0 & k >= v.length or null array or empty array
will result in return value Double.NaN
double r = Double.NaN;
k--; // since arrays are 0-based
if (v!=null && v.length > k && k >= 0) {
Arrays.sort(v);
r = v[k];
}
return r;
| | public static double | median(double[] v)
double r = Double.NaN;
if (v!=null && v.length >= 1) {
int n = v.length;
Arrays.sort(v);
r = (n % 2 == 0)
? (v[n / 2] + v[n / 2 - 1]) / 2
: v[n / 2];
}
return r;
| | public static double | mode(double[] v)if v is zero length or contains no duplicates, return value
is Double.NaN. Else returns the value that occurs most times
and if there is a tie, returns the first such value.
double r = Double.NaN;
// very naive impl, may need to be optimized
if (v!=null && v.length > 1) {
int[] counts = new int[v.length];
Arrays.fill(counts, 1);
for (int i=0, iSize=v.length; i<iSize; i++) {
for (int j=i+1, jSize=v.length; j<jSize; j++) {
if (v[i] == v[j]) counts[i]++;
}
}
double maxv = 0;
int maxc = 0;
for (int i=0, iSize=counts.length; i<iSize; i++) {
if (counts[i] > maxc) {
maxv = v[i];
maxc = counts[i];
}
}
r = (maxc > 1) ? maxv : Double.NaN; // "no-dups" check
}
return r;
| | public static double | stdev(double[] v)
double r = Double.NaN;
if (v!=null && v.length > 1) {
r = Math.sqrt( devsq(v) / (v.length - 1) );
}
return r;
|
|
