Instantiate with Placeholder Values and Generate Sample Data
It’s useful to have a few tools in your back pocket for generating data quickly – either placeholder values (like 1s or 0s) meant to temporarily fill a certain structure you’re instantiating, or sample values that mimic real data, which you can use to explore and experiment with Riptable.
Here’s a brief sampling of Riptable and NumPy methods you can use. For complete details about these functions, see their API reference documentation.
Generate Placeholder Values
The following methods generate repeated 0s and 1s.
10 floating-point zeros:
>>> rt.zeros(10)
FastArray([0., 0., 0., 0., 0., 0., 0., 0., 0., 0.])
10 integer zeros:
>>> rt.zeros(10, int)
FastArray([0, 0, 0, 0, 0, 0, 0, 0, 0, 0])
10 floating-point ones:
>>> rt.ones(10)
FastArray([1., 1., 1., 1., 1., 1., 1., 1., 1., 1.])
10 integer ones:
>>> rt.ones(10, int)
FastArray([1, 1, 1, 1, 1, 1, 1, 1, 1, 1])
The following methods generate repeated specified values.
10 fives:
>>> rt.repeat(5, 10)
FastArray([5, 5, 5, 5, 5, 5, 5, 5, 5, 5])
10 repeats of each array element:
>>> rt.repeat([1, 2], 10)
FastArray([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2])
10 “tiles” of the entire array:
>>> rt.tile([1, 2], 10)
FastArray([1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2])
10 twos:
>>> rt.full(10, 2)
FastArray([2, 2, 2, 2, 2, 2, 2, 2, 2, 2])
Generate Sample Data
Most of these methods generate a range of values.
arange() generates evenly spaced floating-point or integer values
(depending on the input) within a given interval, including the start
value but excluding the stop value. You can also specify a step size
(the spacing between the values; the default is 1). It’s like Python’s
range function, but it returns a FastArray rather than a list.
Numbers 0 through 9:
>>> rt.arange(10)
FastArray([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
Every second number from 1 to 9:
>>> rt.arange(1, 10, 2)
FastArray([1, 3, 5, 7, 9])
For evenly spaced values where the step is a non-integer, it’s better to
use np.linspace. Instead of specifying the step value, you specify
the number of elements. Both the start and stop values are included:
>>> np.linspace(2.0, 3.0, num=5) # Five evenly spaced numbers from 2.0 to 3.0 (the step is 0.25)
array([2. , 2.25, 2.5 , 2.75, 3. ])
For randomly generated values, you have several options.
For integers and floating-point values, NumPy has you covered. Call
default_rng to get a new instance of a NumPy Generator, then call
its methods. To generate values that can be replicated, initialize with
a seed value of your choice to initialize the BitGenerator:
>>> rng = np.random.default_rng(seed=42)
10 floating-point numbers between 0.0 and 1.0 (1.0 excluded):
>>> rng.random(10)
array([0.77395605, 0.43887844, 0.85859792, 0.69736803, 0.09417735,
0.97562235, 0.7611397 , 0.78606431, 0.12811363, 0.45038594])
10 uniformly distributed floats between 0 and 50 (50 excluded):
>>> rng.uniform(0, 50, 10)
array([18.53990121, 46.33824944, 32.193256 , 41.13808066, 22.17070994,
11.36193609, 27.72923935, 3.19086281, 41.3815586 , 31.58321996])
10 integers between 1 and 50 (50 excluded):
>>> rng.integers(1, 50, 10)
array([ 9, 38, 35, 18, 4, 48, 22, 44, 34, 39], dtype=int64)
10 strings chosen from a list:
>>> rng.choice(['GME', 'AMZN', 'TSLA', 'SPY'], 10)
array(['SPY', 'GME', 'AMZN', 'AMZN', 'AMZN', 'GME', 'TSLA', 'GME', 'TSLA', 'TSLA'], dtype='<U4')
10 random Booleans:
>>> rng.choice([True, False], 10)
array([False, False, True, False, True, True, False, True, True, True])
See NumPy’s documentation for more details and other methods.
Riptable has methods for generating random Date and DateTimeNano arrays.
5 DateTimeNanos with NYT time zone:
>>> rt.DateTimeNano.random(5)
DateTimeNano(['20000507 22:02:14.350793900', '20040720 00:24:28.668289697', '19771017 22:34:39.521017110', '20130819 05:29:22.584265022', '20170622 00:50:06.970974486'], to_tz='NYC')
Dates between a start date and an end date (start and end dates included; the default step is 1 day):
>>> rt.Date.range('20190201', '20190208')
Date(['2019-02-01', '2019-02-02', '2019-02-03', '2019-02-04', '2019-02-05', '2019-02-06', '2019-02-07', '2019-02-08'])
5 dates, spaced two days apart, with a specified start date (start date included):
>>> rt.Date.range('20190201', days=5, step=2)
Date(['2019-02-01', '2019-02-03', '2019-02-05', '2019-02-07', '2019-02-09'])
Though Date objects don’t (yet) have a random method, you can
use rng.choice to pick dates from a range:
>>> rt.Date(rng.choice(rt.Date.range('20220201', '20220430'), 5))
Date(['2022-04-12', '2022-02-17', '2022-03-14', '2022-02-12', '2022-04-03'])
Next we cover ways to get data in and out of Riptable: Work with Riptable Files and Other File Formats.
Questions or comments about this guide? Email RiptableDocumentation@sig.com.