brassfibre 0.2.0

Provides multiple-dtype columner storage, known as DataFrame in pandas/R
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134

use std::borrow::Cow;
use std::hash::Hash;

use nullvec::prelude::{Array, Scalar, Nullable, NullVec};
use nullvec::prelude::BasicAggregation as NBasicAggregation;
use nullvec::prelude::NumericAggregation as NNumericAggregation;
use nullvec::prelude::ComparisonAggregation as NComparisonAggregation;


use super::DataFrame;
use indexer::Indexer;
use series::Series;
use traits::{BasicAggregation, NumericAggregation, ComparisonAggregation, Description};

impl<'v, 'i, 'c, I, C> BasicAggregation<'c> for DataFrame<'v, 'i, 'c, I, C>
    where I: Clone + Eq + Hash,
          C: 'c + Clone + Eq + Hash
{
    // ToDo: use 'n lifetime for values
    type Kept = Series<'c, 'c, Scalar, C>;
    type Counted = Series<'c, 'c, usize, C>;

    fn sum(&'c self) -> Self::Kept {
        let ndf = self.get_numeric_data();
        // ToDo: FIXME
        let new_values: Vec<Scalar> = ndf.values.iter().map(|x| x.sum()).collect();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }

    fn count(&'c self) -> Self::Counted {
        let ndf = self.get_numeric_data();
        let new_values: Vec<usize> = ndf.values.iter().map(|x| x.count()).collect();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }
}

impl<'v, 'i, 'c, I, C> NumericAggregation<'c> for DataFrame<'v, 'i, 'c, I, C>
    where I: Clone + Eq + Hash,
          C: 'c + Clone + Eq + Hash
{
    // ToDo: use 'n lifetime for values
    type Coerced = Series<'c, 'c, f64, C>;

    fn mean(&'c self) -> Self::Coerced {
        let ndf = self.get_numeric_data();
        // ToDo: FIXME
        let new_values_tmp: NullVec<f64> = ndf.values.iter().map(|x| x.mean()).collect();
        let new_values: Vec<f64> = new_values_tmp.into();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }

    fn var(&'c self) -> Self::Coerced {
        let ndf = self.get_numeric_data();
        // ToDo: FIXME
        let new_values_tmp: NullVec<f64> = ndf.values.iter().map(|x| x.var()).collect();
        let new_values: Vec<f64> = new_values_tmp.into();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }

    fn unbiased_var(&'c self) -> Self::Coerced {
        let ndf = self.get_numeric_data();
        // ToDo: FIXME
        let new_values_tmp: NullVec<f64> = ndf.values.iter().map(|x| x.unbiased_var()).collect();
        let new_values: Vec<f64> = new_values_tmp.into();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }

    fn std(&'c self) -> Self::Coerced {
        let ndf = self.get_numeric_data();
        // ToDo: FIXME
        let new_values_tmp: NullVec<f64> = ndf.values.iter().map(|x| x.std()).collect();
        let new_values: Vec<f64> = new_values_tmp.into();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }

    fn unbiased_std(&'c self) -> Self::Coerced {
        let ndf = self.get_numeric_data();
        // ToDo: FIXME
        let new_values_tmp: NullVec<f64> = ndf.values.iter().map(|x| x.unbiased_std()).collect();
        let new_values: Vec<f64> = new_values_tmp.into();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }
}

impl<'v, 'i, 'c, I, C> ComparisonAggregation<'c> for DataFrame<'v, 'i, 'c, I, C>
    where I: Clone + Eq + Hash,
          C: 'c + Clone + Eq + Hash
{
    // ToDo: use 'n lifetime for values
    type Kept = Series<'c, 'c, Scalar, C>;

    fn min(&'c self) -> Self::Kept {
        let ndf = self.get_numeric_data();
        // ToDo: FIXME
        let new_values: Vec<Scalar> = ndf.values.iter().map(|x| x.min()).collect();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }

    fn max(&'c self) -> Self::Kept {
        let ndf = self.get_numeric_data();
        // ToDo: FIXME
        let new_values: Vec<Scalar> = ndf.values.iter().map(|x| x.max()).collect();
        Series::from_cow(Cow::Owned(new_values), ndf.columns)
    }
}

impl<'v, 'i, 'c, I, C> Description<'c> for DataFrame<'v, 'i, 'c, I, C>
    where I: Clone + Eq + Hash,
          C: Clone + Eq + Hash
{
    type Described = DataFrame<'v, 'c, 'c, &'c str, C>;

    fn describe(&'c self) -> Self::Described {
        let ndf = self.get_numeric_data();

        let new_index: Vec<&str> = vec!["count", "mean", "std", "min", "max"];

        let describe = |x: &Array| {
            let values: Vec<Nullable<f64>> = vec![Nullable::new(x.count() as f64),
                                                  x.mean(),
                                                  x.std(),
                                                  x.min().as_f64(),
                                                  x.max().as_f64()];
            let nvalues: NullVec<f64> = values.into();
            Array::Float64Array(nvalues)
        };

        let new_values: Vec<Cow<Array>> = ndf.values
            .iter()
            .map(|ref x| Cow::Owned(describe(x)))
            .collect();
        DataFrame::from_cow(new_values, Cow::Owned(Indexer::new(new_index)), ndf.columns)
    }
}