https://sourceforge.net/p/pymdfparser/blog/2013-06-28T05:02:46.470000ZRecent posts to blog2013-06-28T05:02:46.470000Z2013-06-28T05:02:46.470000Zbowiehttps://sourceforge.net/u/bowie2/https://sourceforge.neta5ddd9863145885d4866314678864351e7313674

<div class="markdown_content"><p>Hello,</p> <p>at the moment I am working on the interpolation feature of pyMdfParser.<br /> With feature it is possible to perfomr calculations with signals that were measured in different sample times.</p> <p>I have implemented the following concept:</p> <p>Added a module mdf_operations with a class MDFOps and an interpolation method "interp".<br /> Interface of "interp": A signal list and a desired interpolation rate.<br /> "interp" determines the latest starting time points and the earliest stop time of all time vectors of the signals given in the signal list.</p> <p>Then a new time vector wit the given sample rate is generated between these time boundaries.</p> <p>Then the data vector of all signals given in signal list are interpolated to this new time vector.</p> <p>Interpolation is done by bisectioning followed by a simple linear interpolation.</p> <p>For each interpolated signal a CNBlock object is instantiated and added to the cn_blocks attribute of a CGBlock object.<br /> This CGBlock object is returned to the caller of "interp".</p> <p>In order to have a good usablilty the CGBlock class had to be enhanced:</p> <ul> <li>added a method add --&gt; interface is a CNBlock objects</li> <li>added a method get --&gt; interfcae is a signal name</li> <li>added an attribute hash_table for a fast finding of the signals</li> </ul> <p>The "add" methods does everything to add a CNBlock object to a CGBlock object.<br /> E.g. build the hashtable append the signal name to a signal list.</p> <p>get wirks in the same way as the get methods of MDFParser class but on CGBlock level.</p> <p>As usual I am open for hints a improvements ....</p> <p>Regs </p> <p>Marcus</p></div>

2013-05-04T08:36:46.371000Z2013-05-04T08:36:46.371000Zbowiehttps://sourceforge.net/u/bowie2/https://sourceforge.net5d0e7d8e3d921ff0121ab7cd3388bc7c3aa5fe0a

<div class="markdown_content"><p>Hello,</p> <p>I was busy with some other stuff so I could not add new content the last weeks.</p> <p>With 0.0.6 I have added the "more fancy" info block display and some smaller improvements in the documentation.</p> <p>I think the next big step will be the interpolation to one time channel.</p> <h2 id="concept">Concept:</h2> <h3 id="deal-with-different-start-times">Deal with different start times</h3> <p>It can be that there are completly different start times in the different time vectors.<br /> Worst case is if you have some data acquisition that depends e.g. on a running engine mixed with time triggered signals and you have started the measurement bfore starting the engine.</p> <p>For those cases I will try to find the "earliest" time point which will be the start time point for the commen time vector.<br /> I will fill up missing values of channels with a "later" starting time points with the first value of that respective channels.<br /> At the end of the measure time I will do the same thing in the other direction.<br /> Find "latest" time point which will be the last time point of the common time channel.<br /> Fill up missing values of channels with an earlier last time point with the last values of the respective channels.</p> <p>If someone has a better idea feel free to contribute ...</p></div>

2013-03-27T05:24:36.881000Z2013-03-27T05:24:36.881000Zbowiehttps://sourceforge.net/u/bowie2/https://sourceforge.netcc93ddd01e6f4bab269304b1c563040ab51adc3b

<div class="markdown_content"><p>27.03.2013</p> <p>The problems with the documentation and running the smoe_test from IDLE are solved.<br /> It was just a wrong import operation ...</p> <p>I have fixed the problem and added a documentation in doc\html.zip</p> <p>Next Steps:<br /> - work on documentation (class hierachy diagram)<br /> - add missing conversion types<br /> - add some features like<br /> - interpolation to one time channel<br /> - get info block in a more fancy way<br /> - ....<br /> - test, test, test</p> <p>If you have some ideas for features feel free to bring them in here ...</p> <p>Still looking for more test data.</p></div>

2013-03-24T08:30:18.464000Z2013-03-24T08:30:18.464000Zbowiehttps://sourceforge.net/u/bowie2/https://sourceforge.net89f5bd8220f56bc22ea62264af542e4a4e21346f

<div class="markdown_content"><p>24.03.2013<br /> At the moment I am working on a documentation for pyMdfParser.<br /> I use sphinx and follow the numpy/scipy regulations for documentation.<br /> At the moment I have problems with autodoc and importing mdf_parser for automodule docs. <br /> As soon as the problem is fixed I will add the documentation to the release files ...<br /> It seems that there is a general problem with importing pyMdfParser modules.</p> <p>I must admit that I tested my stuff only in PyDev environment so far and never checked if it can be used with a e.g. IDLE.<br /> I have to fix this first and then come back to the documentation ...</p></div>

2013-03-24T07:05:21.766000Z2013-03-24T07:05:21.766000Zbowiehttps://sourceforge.net/u/bowie2/https://sourceforge.net250d5f06b672455509d240e6b3b6d3dcf61b7903

<div class="markdown_content"><p>Hello,</p> <p>I have decided to write down all my steps and problems of the development in a dev blog<br /> So perhaps if someone wants to join the status of the project should be clear ...</p></div>

2013-03-10T06:38:38.590000Z2013-03-10T06:38:38.590000Zbowiehttps://sourceforge.net/u/bowie2/https://sourceforge.net07f1bbacff0438c2a00e1b8c833eccc6647e80d4

<div class="markdown_content"><p>Hello together,</p> <p>I have added the first draft of a parser for mdf files.<br /> I am looking for people to join ...</p> <p>The aim is, to implement an open source parser in order to use it in Python based applications <br /> which work on mdf measurement data.</p> <p>My problems at the moment:<br /> <em> you only get the mdf 3.3 specification for free in the internet (see Vector or ETAS homepages). It seems there is a new standard (according to the test installation of CANape, where you can select mdf 4 format)<br /> </em> mdf 4 spec is not freely available so we start with mdf 3.3 ...<br /> * real word test data is a problem --&gt; if someone could provide such data ...</p></div>