Gmean, Hmean expanded for other datatypes

integration_tests/test_statistics.py

@@ -1,5 +1,5 @@
 from statistics import (mean, fmean, geometric_mean, harmonic_mean,
-                        variance, stdev, covariance, correlation)
+                        variance, stdev)
 from ltypes import i32, f64, i64
 
 eps: f64
@@ -45,13 +45,31 @@ def test_geometric_mean():
     k = geometric_mean(c)
     assert abs(k - 1.8171205928321397) < eps
 
+    d: list[f64]
+    d = [1.1, 3.4, 17.982, 11.8]
+    l: f64
+    l = geometric_mean(d)
+    assert abs(l - 5.307596520524432) < eps
+
 def test_harmonic_mean():
     c: list[i32]
     c = [9,2,46]
     k: f64
     k = harmonic_mean(c)
     assert abs(k - 4.740458015267175) < eps
 
+    d: list[i32]
+    d = [9, 0, 46]
+    l: f64
+    l = harmonic_mean(d)
+    assert l == 0.0
+
+    e: list[f64]
+    e = [1.1, 3.4, 17.982, 11.8]
+    f: f64
+    f = harmonic_mean(e)
+    assert abs(f - 2.977152988015106) < eps
+
 
 def test_variance():
     a: list[i32]
@@ -66,47 +84,6 @@ def test_variance():
     k = variance(b)
     assert abs(k - 0.40924) < eps
 
-def test_covariance():
-    a: list[i32]
-    a = [1, 2, 3, 4, 5, 6, 7, 8, 9]
-    b: list[i32]
-    b = [1, 2, 3, 1, 2, 3, 1, 2, 3]
-    j: f64
-    j = covariance(a,b)
-    assert abs(j - 0.75) < eps
-
-    c: list[f64]
-    c = [2.74, 1.23, 2.63, 2.22, 3.0, 1.98]
-    d: list[f64]
-    d = [9.4, 1.23, 2.63, 22.4, 1.9, 13.98]
-    k: f64
-    k = covariance(c,d)
-    assert abs(k + 0.24955999999999934) < eps
-
-def test_correlation():
-    a: list[i32]
-    a = [11, 2, 7, 4, 15, 6, 10, 8, 9, 1, 11, 5, 13, 6, 15]
-    b: list[i32]
-    b = [2, 5, 17, 6, 10, 8, 13, 4, 6, 9, 11, 2, 5, 4, 7]
-
-    j: f64
-    j = correlation(a,b)
-    assert abs(j - 0.11521487988958108) < eps
-
-    c: list[i32]
-    c = [1, 2, 3, 4, 5, 6, 7, 8, 9]
-    d: list[i32]
-    d = [9, 8, 7, 6, 5, 4, 3, 2, 1]
-
-    k: f64
-    k = correlation(c,c)
-    assert k == 1.0
-
-    l: f64
-    l = correlation(c,d)
-    assert l == -1.0
-
-
 def test_stdev():
     a: list[i32]
     a = [1, 2, 3, 4, 5]
@@ -128,7 +105,5 @@ def check():
     test_fmean()
     test_variance()
     test_stdev()
-    test_covariance()
-    test_correlation()
 
 check()

src/runtime/statistics.py

@@ -102,7 +102,7 @@ def fmean(x: list[f32]) -> f64:
     """
     return mean(x)
 
-
+@overload
 def geometric_mean(x: list[i32]) -> f64:
     """
     Returns the geometric mean of a data sequence of numbers
@@ -115,11 +115,51 @@ def geometric_mean(x: list[i32]) -> f64:
     i: i32
 
     for i in range(k):
+        if x[i] <= 0:
+            raise Exception("geometric mean requires a non-empty dataset containing positive numbers")
+        product *= float(x[i])
+
+    return product**(1/k)
+
+@overload
+def geometric_mean(x: list[i64]) -> f64:
+    """
+    Returns the geometric mean of a data sequence of numbers
+    """
+    k: i32 = len(x)
+    if k == 0:
+        return 0.0
+    product: f64
+    product = 1.0
+    i: i32
+
+    for i in range(k):
+        if x[i] <= 0:
+            raise Exception("geometric mean requires a non-empty dataset containing positive numbers")
         product *= float(x[i])
 
     return product ** (1 / k)
 
+@overload
+def geometric_mean(x: list[f64]) -> f64:
+    """
+    Returns the geometric mean of a data sequence of numbers
+    """
+    k: i32 = len(x)
+    if k == 0:
+        return 0.0
+    product: f64
+    product = 1.0
+    i: i32
+
+    for i in range(k):
+        if x[i] <= 0.0:
+            raise Exception("geometric mean requires a non-empty dataset containing positive numbers")
+        product *= x[i]
+
+    return product**(1/k)
 
+@overload
 def harmonic_mean(x: list[i32]) -> f64:
     """
     Returns the harmonic mean of a data sequence of numbers
@@ -134,6 +174,49 @@ def harmonic_mean(x: list[i32]) -> f64:
     for i in range(k):
         if x[i] == 0:
             return 0.0
+        if x[i] < 0.0:
+            raise Exception("Harmonic mean does not support negative values")
+        sum += 1 / x[i]
+
+    return float(k/sum)
+
+@overload
+def harmonic_mean(x: list[i64]) -> f64:
+    """
+    Returns the harmonic mean of a data sequence of numbers
+    """
+    k: i32 = len(x)
+    if k == 0:
+        return 0.0
+    sum: f64
+    sum = 0.0
+    i: i32
+
+    for i in range(k):
+        if x[i] == 0:
+            return 0.0
+        if x[i] < 0 :
+            raise Exception("Harmonic mean does not support negative values")
+        sum += 1 / x[i]
+    return k/sum
+
+@overload
+def harmonic_mean(x: list[f64]) -> f64:
+    """
+    Returns the harmonic mean of a data sequence of numbers
+    """
+    k: i32 = len(x)
+    if k == 0:
+        return 0.0
+    sum: f64
+    sum = 0.0
+    i: i32
+
+    for i in range(k):
+        if x[i] == 0.0:
+            return 0.0
+        if x[i] < 0.0:
+            raise Exception("Harmonic mean does not support negative values")
         sum += 1 / x[i]
 
     return k / sum
@@ -189,103 +272,3 @@ def stdev(x: list[i32]) -> f64:
     """
     return variance(x)**0.5
 
-@overload
-def covariance(x: list[i32], y: list[i32]) -> f64:
-    """
-    Returns the covariance of a data sequence of numbers
-    """
-    n: i32 = len(x)
-    m: i32 = len(y)
-    if (n < 2 or m < 2) or n != m:
-        raise Exception("Both inputs must be of the same length (no less than two)")
-    xmean: f64 = mean(x)
-    ymean: f64 = mean(y)
-    num: f64
-    num = 0.0
-    i: i32
-    for i in range(n):
-        num += (x[i] - xmean) * (y[i] - ymean)
-    return num / (n-1)
-
-@overload
-def covariance(x: list[f64], y: list[f64]) -> f64:
-    """
-    Returns the covariance of a data sequence of numbers
-    """
-    n: i32 = len(x)
-    m: i32 = len(y)
-    if (n < 2 or m < 2) or n != m:
-        raise Exception("Both inputs must be of the same length (no less than two)")
-    xmean: f64 = mean(x)
-    ymean: f64 = mean(y)
-    num: f64
-    num = 0.0
-    i: i32
-    for i in range(n):
-        num += (x[i] - xmean) * (y[i] - ymean)
-    return num / (n-1)
-
-@overload
-def correlation(x: list[i32], y: list[i32]) -> f64:
-    """
-    Return the Pearson's correlation coefficient for two inputs.
-    """
-    n: i32 = len(x)
-    m: i32 = len(y)
-    if n != m:
-        raise Exception("correlation requires that both inputs have same number of data points")
-    if n < 2:
-        raise Exception("correlation requires at least two data points")
-    xmean: f64 = mean(x)
-    ymean: f64 = mean(y)
-
-    sxy: f64 = 0.0
-    i: i32
-    for i in range(n):
-        sxy += (x[i] - xmean) * (y[i] - ymean)
-
-    sxx: f64 = 0.0
-    j: i32
-    for j in range(n):
-        sxx += (x[j] - xmean) ** 2
-
-    syy: f64 = 0.0
-    k: i32
-    for k in range(n):
-        syy += (y[k] - ymean) ** 2
-    if sqrt(sxx * syy) == 0:
-        raise Exception('at least one of the inputs is constant')
-    return sxy / sqrt(sxx * syy)
-
-@overload
-def correlation(x: list[f64], y: list[f64]) -> f64:
-    """
-    Return the Pearson's correlation coefficient for two inputs.
-    """
-    n: i32 = len(x)
-    m: i32 = len(y)
-    if n != m:
-        raise Exception("correlation requires that both inputs have same number of data points")
-    if n < 2:
-        raise Exception("correlation requires at least two data points")
-    xmean: f64 = mean(x)
-    ymean: f64 = mean(y)
-
-    sxy: f64 = 0.0
-    i: i32
-    for i in range(n):
-        sxy += (x[i] - xmean) * (y[i] - ymean)
-
-    sxx: f64 = 0.0
-    j: i32
-    for j in range(n):
-        sxx += (x[j] - xmean) ** 2
-
-    syy: f64 = 0.0
-    k: i32
-    for k in range(n):
-        syy += (y[k] - ymean) ** 2
-    if sqrt(sxx * syy) == 0:
-        raise Exception('at least one of the inputs is constant')
-    return sxy / sqrt(sxx * syy)
-