# Consuming all inputs, small outputs **URL:** \ **Category:** GNU Radio\ **Created:** [January 19, 2015, 9:25pm UTC](https://www.ruby-forum.com/t/consuming-all-inputs-small-outputs/233883 "2015-01-19T21:25:37Z")\ **Posts on this page:** 5\ **Page:** 1
**Author:** ![addis\_a](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/a/c4cdca/32.png) [@addis\_a](https://www.ruby-forum.com/u/addis_a)\ **Post date:** [January 19, 2015, 9:25pm UTC](https://www.ruby-forum.com/t/consuming-all-inputs-small-outputs/233883/1 "2015-01-19T21:25:37Z")
I am trying to create a block that outputs one number from each chucnk of data it takes in, I would like to output this number and be able to use it in visualizations, what would the best block be for this? Would a decimation block work? what if I would like to out put three numbers from one chunk of input and visualize those (lets say I’m doing an autocorellation and would like to know sample number I found the highest correlation and the size of the correlation peak) and I want to plot them. Thanks, I’m finding this all REALLY confusing and every time I try something I am just running in to problems ---
**Author:** ![jon\_w](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/3e96dc/32.png) [@jon\_w](https://www.ruby-forum.com/u/jon_w)\ **Post date:** [January 19, 2015, 9:40pm UTC](https://www.ruby-forum.com/t/consuming-all-inputs-small-outputs/233883/2 "2015-01-19T21:40:16Z")
> I am trying to create a block that outputs one number from each chucnk of > data it takes in, I would like to output this number and be able to use it > in visualizations, what would the best block be for this? Would a decimation > block work? what if I would like to out put three numbers from one chunk of > input and visualize those (lets say I’m doing an autocorellation and would > like to know sample number I found the highest correlation and the size of > the correlation peak) and I want to plot them. Thanks, I’m finding this all > REALLY confusing and every time I try something I am just running in to > problems If the output is not a “flow”, then you probably want to use a message port and not a normal port. Normal ports are really meant for streaming sample data and they’re optimized for this with among other things buffering and so if you only produce a very few output, those might end up being buffered for a very long time before being given to the next block. Another option is sometime not to ‘output’ anything, but just have the sample stream pass through your block and just ‘tag’ some samples. The corr\_and\_sync block does that. It searches for a training sequence in an input stream and tags where it find one. So the next block in the chain can take this input stream and know that at this particular sample, there was a correlation peak and it had this value and estimated frequency error. (note that this is correlation with a training sequence, not autocorrelation so might not be applicable to your case). Cheers, Sylvain ---
**Author:** ![jon\_w](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/3e96dc/32.png) [@jon\_w](https://www.ruby-forum.com/u/jon_w)\ **Post date:** [January 19, 2015, 9:40pm UTC](https://www.ruby-forum.com/t/consuming-all-inputs-small-outputs/233883/3 "2015-01-19T21:40:54Z")
Hi Jon, On 01/19/2015 09:24 PM, Jon W. wrote: > I am trying to create a block that outputs one number from each chucnk > of data it takes in, I would like to output this number and be able to > use it in visualizations, what would the best block be for this? > Sounds like a “keep 1 in N” situation, so maybe that; documentation at > [1]. > Would a decimation block work? > Yes. you could for example use a decimating FIR filter (if you don’t > want to write a block of your own), and just set the taps to [1.0], if > you just want to keep the numbers as they were in the original sample > stream. Of course, if you want to do decimation, you’d just try to use > an appropriate low pass filter. > what if I would like to out put three numbers from one chunk of input > and visualize those (lets say I’m doing an autocorellation and would > like to know sample number I found the highest correlation and the > size of the correlation peak) and I want to plot them. > Well, [2] is “keep M in N”. > Thanks, I’m finding this all REALLY confusing and every time I try > something I am just running in to problems. > Hope you have your little moments of success in between 🙂 Greetings, Marcus [1] > **[GNU Radio Manual and C++ API Reference: gr::blocks::keep\_one\_in\_n Class...](https://www.gnuradio.org/doc/doxygen/classgr_1_1blocks_1_1keep __one__ in__n.html)** [2] > **[GNU Radio Manual and C++ API Reference: gr::blocks::keep\_m\_in\_n Class Reference](https://www.gnuradio.org/doc/doxygen/classgr_1_1blocks_1_1keep __m__ in__n.html)** ---
**Author:** ![jon\_w](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/3e96dc/32.png) [@jon\_w](https://www.ruby-forum.com/u/jon_w)\ **Post date:** [January 19, 2015, 10:11pm UTC](https://www.ruby-forum.com/t/consuming-all-inputs-small-outputs/233883/4 "2015-01-19T22:11:28Z")
Hi Jon, um, which tutorial block are you referring to? noutput\_items[0] is the _maximum_ number of items your work function call is allowed to produce (on output stream 0); you can produce less. ninput\_items[0] is the number of items ready for processing (on input stream 0); you can consume less, and will be presented with the remainder plus items that might have come in in the meantime. So the point here is that general\_work of general\_block doesn’t make any assumptions on the relationship between the number of items you consume and produce; if you have a sync\_block, for comparison, there’s always a take-N-give-N relationship. Greetings, Marcus ---
**Author:** ![jon\_w](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/3e96dc/32.png) [@jon\_w](https://www.ruby-forum.com/u/jon_w)\ **Post date:** [January 19, 2015, 10:06pm UTC](https://www.ruby-forum.com/t/consuming-all-inputs-small-outputs/233883/5 "2015-01-19T22:06:23Z")
Well in regards to the general block, in my attempts at learning, I noticed that noutput\_items is less than ninput\_items, what happens to the samples not used? I just ask because the tutorial block is a general block On Mon, Jan 19, 2015 at 3:39 PM, Marcus Müller [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid)