core

BFrame

BFrame

source

BFrame

def BFrame(
    *values, keys:NoneType=None, names:NoneType=None, **fields
):

A native BPL value with shape, keys and names. An array that holds a function as a direct item is a FunctionArray.

test_frame = BFrame(dict(col=[1,2,3]))
test_frame
┌→────────────┐
│ col:┌→────┐ │
│     │1 2 3│ │
│     └ₓ────┘ │
└∊────────────┘
test_eq(BFrame, type(test_frame))

Reading and writing

CSV

source

BFrame.read_csv

def read_csv(
    path, **kwargs
):
bf = BFrame.read_csv("../data/toy.csv", encoding="UTF-8")
bf.df.head(3)
Number City Gender Age Income Illness
0 1 Dallas Male 41 40367.0 No
1 2 Dallas Male 54 45084.0 No
2 3 Dallas Male 42 52483.0 No

source

BFrame.save_csv

def save_csv(
    path
):
# bf.save_csv("../data/save_test.csv");

Selection

source

BFrame.take

def take(
    n:int
):
head = bf.take(100)
head.take(2).df
Number City Gender Age Income Illness
0 1 Dallas Male 41 40367.0 No
1 2 Dallas Male 54 45084.0 No

source

BFrame.head

def head(
    n:int=5
):
bf.head().df
Number City Gender Age Income Illness
0 1 Dallas Male 41 40367.0 No
1 2 Dallas Male 54 45084.0 No
2 3 Dallas Male 42 52483.0 No
3 4 Dallas Male 40 40941.0 No
4 5 Dallas Male 46 50289.0 No

source

BFrame.tail

def tail(
    n:int=5
):
bf.tail().df
Number City Gender Age Income Illness
0 149996 Austin Male 48 93669.0 No
1 149997 Austin Male 25 96748.0 No
2 149998 Austin Male 26 111885.0 No
3 149999 Austin Male 25 111878.0 No
4 150000 Austin Female 37 87251.0 No

Agg

MEAN = add.reduce_first / tally
MEAN(bf.head(100)["Income"])
44362.88

Groupby

fanout
↣(□, {f←↑⍺⋄f ⍵}⍤[0 ∞]⍤[∞ 2])

source

BFrame.matrix

def matrix(
    cols
):

source

BFrame.groupby

def groupby(
    by, cols, **aggs
):
RMSE = MEAN.under(square)
fn_dict=dict(mean=MEAN, rmse=RMSE)
fn_dict
{'mean': +⌿÷≢, 'rmse': (+⌿÷≢)@⊼}
head.groupby("Gender", "Income", rmse=RMSE)
┌→───────────────┐
↓"Male"⌷         │
↓          Income│
│rmse 49632.10639│
│                │
│"Female"⌷       │
│          Income│
│rmse 40738.65853│
└~───────────────┘
head.groupby("Gender", ["Age", "Income"], **fn_dict)
┌→───────────────────────────┐
↓"Male"⌷                     │
↓             Age      Income│
│mean 44.57407407 48301.42593│
│rmse 45.66970793 49632.10639│
│                            │
│"Female"⌷                   │
│             Age      Income│
│mean 45.30434783 39739.36957│
│rmse 46.73468128 40738.65853│
└~───────────────────────────┘

Rolling

windows docs with arg explanations: https://answerdotai.github.io/basedpl/glyphs/up-down-arrow.html

source

BFrame.rolling

def rolling(
    cols, size, edge:str='fill', step:int=1, fill:int=0, **aggs
):
ts = BFrame(dict(price=[5,4.3,4.1,2,10,20,15,25]))
ts.rolling("price", size=3, mean=MEAN, rmse=RMSE).df
price
0 mean 3.100000
rmse 3.807449
1 mean 4.466667
rmse 4.483302
2 mean 3.466667
rmse 3.619392
3 mean 5.366667
rmse 6.345865
4 mean 10.666667
rmse 12.961481
5 mean 15.000000
rmse 15.545632
6 mean 20.000000
rmse 20.412415
7 mean 13.333333
rmse 16.832508 
ts.rolling("price", size=4, step=2, mean=MEAN, rmse=RMSE).df
price
0 mean 3.350000
rmse 3.882654
1 mean 5.100000
rmse 5.901271
2 mean 11.750000
rmse 13.500000
3 mean 15.000000
rmse 17.677670

Viz

source

BFrame.plot

def plot(
    **kwargs
):
BFrame(dict(price=[5,4.3,4.1,2,10,20,15,25])).plot(mark="line", title="Price over time")

source

BFrame.hist_input

def hist_input(
    n:int
):
income = head["Income"]
income.hist_input(5).df
x count
0 21417.9 6
1 31917.7 21
2 42417.5 27
3 52917.3 36
4 63417.1 10

source

BFrame.hist

def hist(
    n:int=50, **kwargs
):
bf["Income"].hist()