Knime Time series

I am very new to knime and have a FFT frequency of a particular Node . To make the time difference linear , I have a day into four group(6 hours each) and now i would like to insert the missing rows . For example see below.

I have tried generating time series like

		<p>Date and time &nbsp; &nbsp; &nbsp;Group</p>

		<p>2012-02-24 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; 0</p>

		<p>2012-02-24 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;1</p>

		<p>2012-02-24 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;2</p>

		<p>2012-02-24 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;3</p>

		<p>2012-02-25 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;0</p>

		<p>2012-02-25 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;1</p>

		<p>2012-02-25 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;2</p>

		<p>2012-02-25 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;3</p>

		<p>And I have a time series like</p>

		<p>Date and time Group</p>

		<p>24.2.2012 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;1</p>

		<p>24.2.2012 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;2</p>

		<p>24.2.2012 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;3</p>

		<p>25.2.2012 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; 0</p>

		<p>25.2.2012 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; 1</p>

		<p>25.2.2012 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;2</p>

		<p>25.2.2012 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; 3</p>

		<p>May i know how to merge them in knime to achieve the below.</p>

		<p>Date and time Group</p>

		<p>2012-02-24 &nbsp; &nbsp; &nbsp;Null</p>

		<p>2012-02-24 &nbsp; &nbsp; &nbsp; &nbsp; 1</p>

		<p>2012-02-24 &nbsp; &nbsp; &nbsp; &nbsp; 2</p>

		<p>2012-02-24 &nbsp; &nbsp; &nbsp; &nbsp; 3</p>

		<p>2012-02-25 &nbsp; &nbsp; &nbsp; &nbsp; 0</p>

		<p>2012-02-25 &nbsp; &nbsp; &nbsp; &nbsp; 1</p>

		<p>2012-02-25 &nbsp; &nbsp; &nbsp; &nbsp; 2</p>

		<p>2012-02-25 &nbsp; &nbsp; &nbsp; &nbsp; 3</p>

		<p>&nbsp;</p>

		<p>&nbsp;</p>
		</td>
		<td style="vertical-align:top">&nbsp;</td>
	</tr>
</tbody>

 

Hello dannychristy,

Is the attached workflow what you are looking for?

Best,
Ferry

Thanks a ton Ferry! It worked!!

 

Cheers\Danny

Hi everybody,

I am new to KNIME but I am impressed by its capabilities already.

I would like to start using Google Trends data in KNIME but I need a little help.  I would like to have access to any nodes, workflows, and even a way to bring Google Trends data automatically.  I know there is a Google Trends "pseudo"  CSV API  (details here:  http://techslides.com/hacking-the-google-trends-api   ) that allows you to collect a limit of 1000 calls per hour, free of cost, but I have no idea how to install this into KNIME.

ANY help you could provide or pre-built nodes / workflows you have will be inmensely appreciated.

Thank you and Best regards,

 

FJ

 

Hi,

 

I am new to KNIME and it is interesting. I have the Frequency and amplitude data of a Machinery Vibration using  multiple sensors. I am trying to add UCL and LCL as suggested in the time series examples. The workflow isn't so clear to understand it. In the 1st level alarm, I have selected 4 diiferent coulmns to get Mean and Standard deviation but then in the Rule Engine node I am unable to add the following as there is no coulmn by the name UCL and LCL recognized .

                                                    UCL = avg + 2* stddev

                                                     LCL = avg – 2* stddev 

and 

                                    $messure$ < $LCL$ => 1

                                         $messure$ > $UCL$ => 1

                                                 TRUE => 0 

 

 Could you please help in adding UCL and LCL in the work flow ? 

            I'd also like to know if we can detect the faults in a Machine using Knime ? I do not have historical data but then I have data collected using four sensors with the faults imposed onto the machine. My ultimate goal is to do analysis in order to detect the fault and suggest taking an action in the form of sending an email. I'd appreciate if you can help me getting through this issue . Thanks for your help in advance.

 

Thanks and regards,

Lakhan Raj

Dear FJ, dear Lakhan,

Please move your questions to a new post. Then I will gladly answer in more detail.

@FJ: As far as I know there are no workflows about that. There is a Google API Connector, but it does not support Google Trends API yet.

@Lakhan: For aggregations like your "UCL" and "LCL" have a look at the Column Aggregator. For the rules please have a look at the Rule Engine node.
Anomaly detection in machines is a great application for the KNIME Analytics Platform. There are use cases and presentations about how to recognize and even predict the failure of parts and automatically schedule maintainance in time to prevent losses due to unexpected failure.

Best,
Ferry

Thank you so much Ferry.  I will post it from next time.  I am not getting how to add   UCL= avg+2*stddev  and       LCL = avg-2*stddev,  in the column aggregation node as you suggested . I have selected the required columns in settings and in option settings how do I change the aggregation method to the one I mentioned above. How do I change the operations ? Thanks for your help. Kindly answer my question as early as possible. 

                                        

                            

Hello Lakhan,

In the case you have two columns with numeric values, like in the example, you can use the Math Formula to perform any formula you like on those two columns (see the example you mentioned EXAMPLES ->060_Webinars ->Whats_New_in_2.10->3 Time Series->Time Series).

About getting a new column: Most nodes that are part of the Data Manipulation category offer you the choice between replacing an existing or appending a new column.
The Math Formula offers this choice as well. If you choose to append the new column you can name it UCL.

I hope that clears it up, if you have further questions please open a new post.

Best,
Ferry