I am trying to do crossover on a Genetic Algorithm population using numpy.
I have sliced the population using parent 1 and parent 2.
population = np.random.randint(2, size=(4,8))
p1 = population[::2]
p2 = population[1::2]
But I am not able to figure out any lambda or numpy command to do a multi-point crossover over parents.
The concept is to take ith row of p1 and randomly swap some bits with ith row of p2.
I think you want to select from p1 and p2 at random, cell by cell.
To make it easier to understand i've changed p1 to be 10 to 15 and p2 to be 20 to 25. p1 and p2 were generated at random in these ranges.
p1
Out[66]:
array([[15, 15, 13, 14, 12, 13, 12, 12],
[14, 11, 11, 10, 12, 12, 10, 12],
[12, 11, 14, 15, 14, 10, 13, 10],
[11, 12, 10, 13, 14, 13, 12, 13]])
In [67]: p2
Out[67]:
array([[23, 25, 24, 21, 24, 20, 24, 25],
[21, 21, 20, 20, 25, 22, 24, 22],
[24, 22, 25, 20, 21, 22, 21, 22],
[22, 20, 21, 22, 25, 23, 22, 21]])
In [68]: sieve=np.random.randint(2, size=(4,8))
In [69]: sieve
Out[69]:
array([[0, 1, 0, 1, 1, 0, 1, 0],
[1, 1, 1, 0, 0, 1, 1, 1],
[0, 1, 1, 0, 0, 1, 1, 0],
[0, 0, 0, 1, 1, 1, 1, 1]])
In [70]: not_sieve=sieve^1 # Complement of sieve
In [71]: pn = p1*sieve + p2*not_sieve
In [72]: pn
Out[72]:
array([[23, 15, 24, 14, 12, 20, 12, 25],
[14, 11, 11, 20, 25, 12, 10, 12],
[24, 11, 14, 20, 21, 10, 13, 22],
[22, 20, 21, 13, 14, 13, 12, 13]])
The numbers in the teens come from p1 when sieve is 1
The numbers in the twenties come from p2 when sieve is 0
This may be able to be made more efficient but is this what you expect as output?
Related
import numpy as np
data = np.array([[10, 20, 30, 40, 50, 60, 70, 80, 90],
[2, 7, 8, 9, 10, 11],
[3, 12, 13, 14, 15, 16],
[4, 3, 4, 5, 6, 7, 10, 12]],dtype=object)
target = data[:,0]
It has this error.
IndexError Traceback (most recent call last)
Input In \[82\], in \<cell line: 9\>()
data = np.array(\[\[10, 20, 30, 40, 50, 60, 70, 80, 90\],
\[2, 7, 8, 9, 10, 11\],
\[3, 12, 13, 14, 15, 16\],
\[4, 3, 4, 5, 6, 7, 10,12\]\],dtype=object)
# Define the target data ----\> 9 target = data\[:,0\]
IndexError: too many indices for array: array is 1-dimensional, but 2 were indexed
May I know how to fix it, please? I mean do not change the elements in the data. Many thanks. I made the matrix in the same size and the error message was gone. But I have the data with variable size.
You have a array of objects, so you can't use indexing on axis=1 as there is none (data.shape -> (4,)).
Use a list comprehension:
out = np.array([a[0] for a in data])
Output: array([10, 2, 3, 4])
I have a dataset like so -
15643, 14087, 12020, 8402, 7875, 3250, 2688, 2654, 2501, 2482, 1246, 1214, 1171, 1165, 1048, 897, 849, 579, 382, 285, 222, 168, 115, 92, 71, 57, 56, 51, 47, 43, 40, 31, 29, 29, 29, 29, 28, 22, 20, 19, 18, 18, 17, 15, 14, 14, 12, 12, 11, 11, 10, 9, 9, 8, 8, 8, 8, 7, 6, 5, 5, 5, 4, 4, 4, 3, 3, 3, 3, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
Based on domain knowledge, I know that larger values are the only ones we want to include in our analysis. How do I determine where to cut off our analysis? Should it be don't include 15 and lower or 50 and lower etc?
You can do a distribution check with quantile function. Then you can remove values below lowest 1 percentile or 2 percentile. Following is an example:
import numpy as np
data = np.array(data)
print(np.quantile(data, (.01, .02)))
Another method is calculating the inter quartile range (IQR) and setting lowest bar for analysis is Q1-1.5*IQR
Q1, Q3 = np.quantile(data, (0.25, 0.75))
data_floor = Q1 - 1.5 * (Q3 - Q1)
I have a numpy array (inputs) of shape (30,1). I want to insert 31st value (eg. x = 2). Trying to use the np.insert function but it is giving me out of bounds error.
np.insert(inputs,b+1,x)
IndexError: index 31 is out of bounds for axis 0 with size 30
Short answer: you need to insert it at index b, not b+1.
The index you pass to np.insert(..) [numpy-doc], is the one where the element should be added. If you insert it at index 30, then it will be positioned last. Note that indexes are zero-based. So if you have an array with 30 elements, then the last index is 29. If you thus insert this at index 30, we get:
>>> a
array([ 0, 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])
>>> np.insert(a,30,42)
array([ 0, 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, 42])
I have a small matrix A with dimensions MxNxO
I have a large matrix B with dimensions KxMxNxP, with P>O
I have a vector ind of indices of dimension Ox1
I want to do:
B[1,:,:,ind] = A
But, the lefthand of my equation
B[1,:,:,ind].shape
is of dimension Ox1xMxN and therefore I can not broadcast A (MxNxO) into it.
Why does accessing B in this way change the dimensions of the left side?
How can I easily achieve my goal?
Thanks
There's a feature, if not a bug, that when slices are mixed in the middle of advanced indexing, the sliced dimensions are put at the end.
Thus for example:
In [204]: B = np.zeros((2,3,4,5),int)
In [205]: ind=[0,1,2,3,4]
In [206]: B[1,:,:,ind].shape
Out[206]: (5, 3, 4)
The 3,4 dimensions have been placed after the ind, 5.
We can get around that by indexing first with 1, and then the rest:
In [207]: B[1][:,:,ind].shape
Out[207]: (3, 4, 5)
In [208]: B[1][:,:,ind] = np.arange(3*4*5).reshape(3,4,5)
In [209]: B[1]
Out[209]:
array([[[ 0, 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]]])
This only works when that first index is a scalar. If it too were a list (or array), we'd get an intermediate copy, and couldn't set the value like this.
https://docs.scipy.org/doc/numpy-1.15.0/reference/arrays.indexing.html#combining-advanced-and-basic-indexing
It's come up in other SO questions, though not recently.
weird result when using both slice indexing and boolean indexing on a 3d array
I keep getting the following error: "Uncaught Requested parser does not exist" from the devtool console
and no chart is displayed on my dashboard.
JS Code:
var chart = new Keen.Dataviz()
.el('#chart-01')
.height(280)
.title('Registered CSRs')
.type('bar')
.prepare();
// Fetch data from my server's API
var json = $.ajax({
url: "/api/v1.0/registered_csrs/nwg",
dataType: "json",
success: function (jsonData) {
chart
.data(jsonData)
.render();
}
});
Here is the format of jsonData
$ curl -GET http://localhost:4000/api/v1.0/registered_csrs/nwg
[["Date", "Total", "Emergency", "High", "Medium", "Low"],
["2016-02-01", 1, 0, 0, 1, 0],
["2016-03-01", 6, 0, 0, 6, 0],
["2016-04-01", 11, 0, 1, 7, 3],
["2016-05-01", 19, 0, 1, 16, 2],
["2016-06-01", 27, 0, 12, 13, 2],
["2016-07-01", 27, 3, 12, 12, 0],
["2016-08-01", 25, 3, 8, 11, 3],
["2016-09-01", 21, 4, 10, 5, 2],
["2016-10-01", 19, 3, 4, 11, 1],
["2016-11-01", 29, 4, 12, 12, 1],
["2016-12-01", 26, 2, 9, 14, 1],
["2017-01-01", 16, 1, 3, 11, 1],
["2017-02-01", 22, 2, 8, 11, 1],
["2017-03-01", 28, 2, 10, 14, 2],
["2017-04-01", 15, 2, 6, 5, 2],
["2017-05-01", 28, 2, 7, 18, 1],
["2017-06-01", 22, 1, 11, 8, 2],
["2017-07-01", 10, 1, 4, 5, 0]]
Take a look
Keen.io Dataviz to draw graph but keep getting error "Uncaught Requested parser does not exist"
https://keen.io/docs/visualize/visualize-your-own-data/
Code below should work fine in your app
chart
.data({result: jsonData})
.render();