<div dir="ltr"><br><br><div class="gmail_quote">---------- Forwarded message ----------<br>From: <b class="gmail_sendername">Thomas Tibbitts</b> <span dir="ltr"><<a href="mailto:Thomas.Tibbitts@infi.com">Thomas.Tibbitts@infi.com</a>></span><br>
Date: 11 June 2014 21:16<br>Subject: Re: [NMF-user] matrix size limit?<br>To: Renaud Gaujoux <<a href="mailto:renaud@cbio.uct.ac.za">renaud@cbio.uct.ac.za</a>><br><br><br>The problem was rows of all zeros in the input matrix.<br>

Thank you for your help.<br>
Best regards,<br>
Tom<br>
<br>
Sent from my iPhone<br>
<div class="HOEnZb"><div class="h5"><br>
> On May 26, 2014, at 4:25 AM, "Renaud Gaujoux" <<a href="mailto:renaud@cbio.uct.ac.za">renaud@cbio.uct.ac.za</a>> wrote:<br>
><br>
> Are you sure you are reading the data correctly? The names seem strange.<br>
><br>
><br>
>> On 23 May 2014 22:28, Thomas Tibbitts <<a href="mailto:Thomas.Tibbitts@infi.com">Thomas.Tibbitts@infi.com</a>> wrote:<br>
>> Hi Renaud,<br>
>><br>
>> Thank you for your quick response.<br>
>><br>
>> This test returns TRUE:<br>
>> min(rowSums(x, na.rm = TRUE), na.rm = TRUE) == 0<br>
>> [1] TRUE<br>
>><br>
>> Then I received the following output with the next two commands:<br>
>> s <- rowSums(x, na.rm = TRUE)<br>
>> str(which(<a href="http://is.na" target="_blank">is.na</a>(s) | s==0))<br>
>> Named int [1:1691] 1 2 3 4 5 6 7 8 9 10 ...<br>
>> - attr(*, "names")= chr [1:1691] "ABCA13_11152_A2223P,0,0,0,0,0,0,0.46,0.5,0.53,0.51,0,0,0,0,0,0,0,0,0.48,0.45,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.47,0.46,0,0,0,0,0.48"| __truncated__ "ABCA13_11157_K2306T,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0"| __truncated__ "ABCA13_11161_S2537A,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.5,0.5,0,0"| __truncated__ "ABCA13_11163_D2630G,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0"| __truncated__ ...<br>

>><br>
>> After the header row, the first few lines of the data in my file are:<br>
>> ABCA13_11152_A2223P,0,0,0,0,0,0,0.46,0.5,0.53,0.51,0,0,0,0,0,0,0,0,0.48,0.45,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.47,0.46,0,0,0,0,0.48,0.48,0,0,0,0,0.47,0.53,0,0,0,0,0,0,0,0,0,0,0,0,0,0.56,0,0,0.5,0.48,0,0,0,0,0,0,0,0,0.48,0.46,0.45,0.47,0,0,0,0,0,0,0,0,0,0,0,0,0,0<br>

>> ABCA13_11157_K2306T,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.48,0.5,0,0,0,0,0,0,0,0,0,0,0.54,0.48,0,0,0,0,0,0,0,0<br>

>> ABCA13_11161_S2537A,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.5,0.5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.48,0.48,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0<br>

>> ABCA13_11163_D2630G,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.47,0.48,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0<br>

>><br>
>> So the error message did not help. Do you have any other ideas?<br>
>> I am not sure how to turn on logging in R to get a complete log.<br>
>><br>
>> Thank you again for your help!<br>
>> Best regards,<br>
>> Tom<br>
>><br>
>> -----Original Message-----<br>
>> From: <a href="mailto:getoxxx@gmail.com">getoxxx@gmail.com</a> [mailto:<a href="mailto:getoxxx@gmail.com">getoxxx@gmail.com</a>] On Behalf Of Renaud Gaujoux<br>
>> Sent: Monday, May 19, 2014 4:58 PM<br>
>> To: Thomas Tibbitts<br>
>> Cc: <a href="mailto:nmf-user@lists.r-forge.r-project.org">nmf-user@lists.r-forge.r-project.org</a><br>
>> Subject: Re: [NMF-user] matrix size limit?<br>
>><br>
>> Hi Thomas,<br>
>><br>
>> this is not really a large matrix and should not pose a problem, for example on a toy dataset:<br>
>><br>
>>> x <- rmatrix(4000, 100)<br>
>>> res <- nmf(x, 3)<br>
>>> runtime(res)<br>
>>   user  system elapsed<br>
>> 19.888   0.004  19.899<br>
>>> res <- nmf(x, 3, nrun = 3)<br>
>>> runtime(res)<br>
>>   user  system elapsed<br>
>> 33.262   0.004  33.447<br>
>><br>
>> Do you see warnings as well? Sending a complete log would help.<br>
>> The error you see is thrown if the following test is TRUE:<br>
>><br>
>> min(rowSums(x, na.rm = TRUE), na.rm = TRUE) == 0<br>
>><br>
>> Can you please directly run this test on your data? It should return TRUE. Then look at:<br>
>><br>
>> s <- rowSums(x, na.rm = TRUE)<br>
>> str(which( <a href="http://is.na" target="_blank">is.na</a>(s) | s == 0 ))<br>
>><br>
>> which will give you a hint on which row(s) are detected as problematic. By the way, is your dataset nonnegative?<br>
>><br>
>> Bests,<br>
>> Renaud<br>
>><br>
>>> On 19 May 2014 18:37, Thomas Tibbitts <<a href="mailto:Thomas.Tibbitts@infi.com">Thomas.Tibbitts@infi.com</a>> wrote:<br>
>>> NMF-users,<br>
>>><br>
>>><br>
>>><br>
>>> I can run NMF with a matrix of 1646 rows x 100 columns without error.<br>
>>><br>
>>><br>
>>><br>
>>> However when I combine data sets into a larger matrix of 3495 rows x<br>
>>> 100 columns I keep getting the error:<br>
>>><br>
>>> Error:  NMF::nmf  – Input matrix x contains at least one null or<br>
>>> NA-filled row.<br>
>>><br>
>>><br>
>>><br>
>>> There are no nulls or NA-filled rows in the larger matrix, I wrote it<br>
>>> out to disk to make sure.<br>
>>><br>
>>><br>
>>><br>
>>> So I ‘m thinking I hit an matrix size limitation.<br>
>>><br>
>>> How can I fix this problem?<br>
>>><br>
>>><br>
>>><br>
>>> Thank you.<br>
>>><br>
>>> Tom<br>
>>><br>
>>> This message and any attachments may contain confidential information<br>
>>> and/or privileged material. If you received this message in error,<br>
>>> please contact the sender and delete this message and any attachments<br>
>>> immediately. Thank you.<br>
>>><br>
>>> _______________________________________________<br>
>>> nmf-user mailing list<br>
>>> <a href="mailto:nmf-user@lists.r-forge.r-project.org">nmf-user@lists.r-forge.r-project.org</a><br>
>>> <a href="https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/nmf-user" target="_blank">https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/nmf-user</a><br>
>><br>
>><br>
>><br>
>> --<br>
>> Renaud Gaujoux, PhD<br>
>> Systems Immunology - Technion, Haifa, Israel<br>
>> This message and any attachments may contain confidential information and/or privileged material. If you received this message in error, please contact the sender and delete this message and any attachments immediately. Thank you.<br>

><br>
><br>
><br>
> --<br>
> Renaud Gaujoux, PhD<br>
> Systems Immunology - Technion, Haifa, Israel<br>
This message and any attachments may contain confidential information and/or privileged material. If you received this message in error, please contact the sender and delete this message and any attachments immediately. Thank you.<br>

<br>
</div></div></div><br><br clear="all"><br>-- <br><div dir="ltr"><div><div><font><span style="font-family:arial,helvetica,sans-serif">Renaud Gaujoux, PhD<br></span></font></div><font><span style="font-family:arial,helvetica,sans-serif">Systems Immunology - Technion, Haifa, Israel<br>
</span></font></div><font><span style="font-family:arial,helvetica,sans-serif"></span></font><div><font><br></font></div></div>
</div>