Xeve - About code and problems

Linearize nonlinear state space equation in MATLAB at steady state setpoint

July 30, 2018 0 comments Article Uncategorized nspo

This post shows one way to linearize a nonlinear state equation $\dot{x} = f(x,u)$ at a steady state setpoint $(x_0, u_0)$ in MATLAB. It is assumed that a function ode.m exists in which the state equation is implemented:

function dx = ode(t, x, u)

 % example ODE
dx1 = tan(x(4)) * x(2) + 2*u(2) - 1;
dx2 = x(1) - sin(x(2));
dx3 = 13 * x(4) + u(1) + 1;
dx4 = 2*x(1);

dx = [dx1; dx2; dx3; dx4];
end

function dx = ode(t, x, u)

% example ODE

dx1 = tan(x(4)) * x(2) + 2*u(2) - 1;

dx2 = x(1) - sin(x(2));

dx3 = 13 * x(4) + u(1) + 1;

dx4 = 2*x(1);

dx = [dx1; dx2; dx3; dx4];

end

Calculate Levenshtein distance in MATLAB (words or characters) with Wagner–Fischer algorithm

July 1, 2018 0 comments Article Uncategorized nspo

function y = lvdistance(sentence1, sentence2, penalties, split_characters)
 %LVDISTANCE Calculate Levenshtein distance between two strings
 % Uses Wagner–Fischer algorithm
 % See https://en.wikipedia.org/wiki/Wagner%E2%80%93Fischer_algorithm
 %
 % Arguments:
 % sentence1, sentence2: strings, e.g. 'hello world'
 % penalties (optional): vector of penalties for deletion, insertion and
 % substitution (default: [1 1 1])
 % split_characters (optional): split at each character, not only at spaces
 % (default: false)
 %
 % GPL v3, Nicolai Spohrer <nicolai@xeve.de>

switch nargin
    case 2
        penalty_deletion = 1;
        penalty_insertion = 1;
        penalty_substitution = 1;
        split_characters = false;
    case 3
        penalty_deletion = penalties(1);
        penalty_insertion = penalties(2);
        penalty_substitution = penalties(3);
        split_characters = false;
    case 4
        penalty_deletion = penalties(1);
        penalty_insertion = penalties(2);
        penalty_substitution = penalties(3);
    otherwise
        error 'Invalid number of arguments';
end

if split_characters
    % split at each character
    sentence1 = num2cell(sentence1);
    sentence2 = num2cell(sentence2);
else
    % split at spaces
    sentence1 = strsplit(sentence1, ' ');
    sentence2 = strsplit(sentence2, ' ');
end



m = length(sentence1);
n = length(sentence2);

 % distance 2D array
d = zeros(m+1, n+1);

for i=1:m+1
    d(i, 1) = i-1;
end

for j=1:n+1
    d(1, j) = j-1;
end

for j=1:n
    for i=1:m
        if strcmp(sentence1{i}, sentence2{j})
            % words are equal, therefore no operation required
            d(i+1, j+1) = d(i, j);
        else
            % words are not equal
            d(i+1, j+1) = min(min(...
                d(i, j+1) + penalty_deletion,...  % deletion
                d(i+1, j) + penalty_insertion),...  % insertion
                d(i, j) + penalty_substitution); % substitution
        end
    end
end

y = d(m+1, n+1);
end

function y = lvdistance(sentence1, sentence2, penalties, split_characters)

%LVDISTANCE Calculate Levenshtein distance between two strings

% Uses Wagner–Fischer algorithm

% See https://en.wikipedia.org/wiki/Wagner%E2%80%93Fischer_algorithm

% Arguments:

% sentence1, sentence2: strings, e.g. 'hello world'

% penalties (optional): vector of penalties for deletion, insertion and

% substitution (default: [1 1 1])

% split_characters (optional): split at each character, not only at spaces

% (default: false)

% GPL v3, Nicolai Spohrer <nicolai@xeve.de>

switch nargin

case 2

penalty_deletion = 1;

penalty_insertion = 1;

penalty_substitution = 1;

split_characters = false;

case 3

penalty_deletion = penalties(1);

penalty_insertion = penalties(2);

penalty_substitution = penalties(3);

split_characters = false;

case 4

penalty_deletion = penalties(1);

penalty_insertion = penalties(2);

penalty_substitution = penalties(3);

otherwise

error 'Invalid number of arguments';

end

if split_characters

% split at each character

sentence1 = num2cell(sentence1);

sentence2 = num2cell(sentence2);

else

% split at spaces

sentence1 = strsplit(sentence1, ' ');

sentence2 = strsplit(sentence2, ' ');

end

m = length(sentence1);

n = length(sentence2);

% distance 2D array

d = zeros(m+1, n+1);

for i=1:m+1

d(i, 1) = i-1;

end

for j=1:n+1

d(1, j) = j-1;

end

for j=1:n

for i=1:m

if strcmp(sentence1{i}, sentence2{j})

% words are equal, therefore no operation required

d(i+1, j+1) = d(i, j);

else

% words are not equal

d(i+1, j+1) = min(min(...

d(i, j+1) + penalty_deletion,... % deletion

d(i+1, j) + penalty_insertion),... % insertion

d(i, j) + penalty_substitution); % substitution

end

y = d(m+1, n+1);

end

Convert dSPACE ControlDesk measurements to MATLAB/Simulink timeseries (updated)

May 29, 2018 4 comments Article Uncategorized nspo

(Old version here)

With this updated MATLAB script, multiple dSPACE ControlDesk measurements can be imported with a single execution
The result is saved as a tscollection (collection of timeseries) and all signals are saved with their corresponding name
Access imported data with Simulink ‘From Workspace’ block and, e.g., dsp_tscs{1}.my_signal

Fix for Latex4CorelDraw with CorelDraw 2017

May 19, 2018 0 comments Article Uncategorized nspo

I was not able to make the official build of Latex4CorelDraw work with CorelDraw 2017 (and Windows 10). The ‘Latex’ Docker window was always empty. The solution was to recompile it with Visual Studio 2017 Community.

Convert dSPACE ControlDesk measurement to MATLAB timeseries that can be used in Simulink

May 11, 2018 1 comment Article Uncategorized nspo

(Updated version here)

dSPACE ControlDesk can export measurements to mdf4 files or mat files
mat file exports can be converted to timeseries with this MATLAB script
The timeseries can then be imported into Simulink with the ‘From Workspace’ blog
Multiple signals can be imported at once

Example output:

>> import_dspace_mat_to_simulink_ts
--Imported dSPACE mat file 
C:\Users\Nicolai\Nextcloud\foo.mat
into workspace variable dsp_ts
--Time: 0.000000 s to 42.399572 s, 423980 steps
--2 variables are in dsp_ts:
 'In1' 'In1_1'

>> import_dspace_mat_to_simulink_ts

--Imported dSPACE mat file

C:\Users\Nicolai\Nextcloud\foo.mat

into workspace variable dsp_ts

--Time: 0.000000 s to 42.399572 s, 423980 steps

--2 variables are in dsp_ts:

'In1' 'In1_1'

Robust sync between Android phone and Linux computer

April 22, 2018 0 comments Article Uncategorized nspo

I already tried quite a few methods to reliably sync data (e.g. backups) between my Android phone and a Linux computer. Most were not good, some were horrible (MTP…). I just tried one that has proven to be the best, by far: installing an SSH server on the phone and then syncing with rsync. I Read More

Django DetailView with pagination for related objects

March 31, 2018 0 comments Article Uncategorized nspo

This is an example of a DetailView of a User that can have a paginated list of Purchases:

from django.views.generic.detail import DetailView
from django.core import paginator

@method_decorator(staff_member_required(login_url='user_login'), name='dispatch')
class UserDetailView(DetailView):
    model = User
    template_name = "user_detail.html"
    purchases_paginate_by = 5

    def get_context_data(self, **kwargs):
        context = super(UserDetailView, self).get_context_data(**kwargs)

        purchases_page = self.request.GET.get("purchases_page")
        purchases = self.object.get_purchases().filter()
        purchases_paginator = paginator.Paginator(purchases, self.purchases_paginate_by)
        # Catch invalid page numbers
        try:
            purchases_page_obj = purchases_paginator.page(purchases_page)
        except (paginator.PageNotAnInteger, paginator.EmptyPage):
            purchases_page_obj = purchases_paginator.page(1)

        context["purchases_page_obj"] = purchases_page_obj
        return context

from django.views.generic.detail import DetailView

from django.core import paginator

@method_decorator(staff_member_required(login_url='user_login'), name='dispatch')

class UserDetailView(DetailView):

model = User

template_name = "user_detail.html"

purchases_paginate_by = 5

def get_context_data(self, **kwargs):

context = super(UserDetailView, self).get_context_data(**kwargs)

purchases_page = self.request.GET.get("purchases_page")

purchases = self.object.get_purchases().filter()

purchases_paginator = paginator.Paginator(purchases, self.purchases_paginate_by)

# Catch invalid page numbers

try:

purchases_page_obj = purchases_paginator.page(purchases_page)

except (paginator.PageNotAnInteger, paginator.EmptyPage):

purchases_page_obj = purchases_paginator.page(1)

context["purchases_page_obj"] = purchases_page_obj

return context

In the template file, the pagination needs to done similar to the following example (using django-bootstrap3):

{% bootstrap_pagination purchases_page_obj extra=request.GET.urlencode parameter_name="purchases_page" %}

1	{% bootstrap_pagination purchases_page_obj extra=request.GET.urlencode parameter_name="purchases_page" %}

Github gist Read More

Export object_list of a django-filters FilterView to CSV

March 31, 2018 0 comments Article Uncategorized nspo

This is a simple way to allow users to export the data they have just filtered using django-filter as CSV.

# Normal View (has a link to export view and keeps GET parameters)
@method_decorator(staff_member_required(login_url='user_login'), name='dispatch')
class PaymentListView(FilterView):
    filterset_class = filters.PaymentFilter
    template_name = "barsys/userarea/payment_list.html"
    paginate_by = 10

# Export View
@method_decorator(staff_member_required(login_url='user_login'), name='dispatch')
class PaymentExportView(FilterView):
    filterset_class = filters.PaymentFilter

    def render_to_response(self, context, **response_kwargs):
        # Could use timezone.now(), but that makes the string much longer
        filename = "{}-export.csv".format(datetime.datetime.now().replace(microsecond=0).isoformat())

        response = HttpResponse(content_type='text/csv')
        response['Content-Disposition'] = 'attachment; filename="{}"'.format(filename)

        writer = csv.writer(response)
        writer.writerow(['created', 'email', 'display_name'])

        for obj in self.object_list:
            writer.writerow([obj.created_date, obj.user.email, obj.user.display_name])

        return response

# Normal View (has a link to export view and keeps GET parameters)

@method_decorator(staff_member_required(login_url='user_login'), name='dispatch')

class PaymentListView(FilterView):

filterset_class = filters.PaymentFilter

template_name = "barsys/userarea/payment_list.html"

paginate_by = 10

# Export View

@method_decorator(staff_member_required(login_url='user_login'), name='dispatch')

class PaymentExportView(FilterView):

filterset_class = filters.PaymentFilter

def render_to_response(self, context, **response_kwargs):

# Could use timezone.now(), but that makes the string much longer

filename = "{}-export.csv".format(datetime.datetime.now().replace(microsecond=0).isoformat())

response = HttpResponse(content_type='text/csv')

response['Content-Disposition'] = 'attachment; filename="{}"'.format(filename)

writer = csv.writer(response)

writer.writerow(['created', 'email', 'display_name'])

for obj in self.object_list:

writer.writerow([obj.created_date, obj.user.email, obj.user.display_name])

return response

Github gist Read More

Running multiple Django projects on one Apache instance with mod_wsgi

March 31, 2018 0 comments Article Uncategorized nspo

I did not yet find a page with a concise but complete description of how to run multiple Django instances/projects on one server with Apache. This method is similar to the one in the Django docs, but the docs version does not run the Django instance in a daemon process, therefore it does not work Read More

Analytical convolution integral (Analytische Faltung) with Matlab and Maple

September 10, 2015 0 comments Article Uncategorized nspo

With $\sigma(t) = Heaviside(t)$ and $\delta(t) = Dirac(t)$
Function 1 (e.g. input signal/Eingangssignal):
$u(t) = \sigma(t-1) - \sigma(t-4)$

Function 2 (e.g. impulse response/Stoßantwort):
$g(t) = -\frac{1}{RC}\exp(-\frac{t}{RC})\sigma(t) + \delta(t)$

1 2 »