kdahlquist / GRNmap

Gene Regulatory Network modeling and parameter estimation
BSD 3-Clause "New" or "Revised" License
4 stars 3 forks source link

Missing data error on v1.10 pre-release (beta) #411

Closed kdahlquist closed 6 years ago

kdahlquist commented 6 years ago

image

image

I'm running the input workbook pasted below on v1.10 pre-release (beta) and am getting the errors shown above. I don't understand what is going on because line 2, which would be CIN5, doesn't have the problem, and if there is a single replicate, I though we were allowing it to run, just not computing sigmas for it. 17-genes_32-edges_db3_Sigmoid_estimation_all-timepoints_no-dHAP4-data.xlsx

kdahlquist commented 6 years ago

These two workbooks give similar errors, but I can't see what's actually wrong with the expression data. 16-genes_27-edges_db6_Sigmoid_estimation_all-timepoints.xlsx 15-genes_34-edges_db7_Sigmoid_estimation_all-timepoints_no-dHAP4-data.xlsx

im-deepfriedwater commented 6 years ago

I'll be taking a look at this!

im-deepfriedwater commented 6 years ago

Preliminary digging is leading me into the dataCell being an empty matrix which is why it's erroring out. This seems to be because the compressMissingData function (or perhaps the convertToNestedStructure function) is creating the compressed data structure incorrectly. Will still have to do more debugging though to find the logical error.

im-deepfriedwater commented 6 years ago

image

Oh.. more digging has lead me to just outright ouput the rawData that was read-in, and I've found a lot of NaNs. Seems to me it's a formatting issue again. Will take a look at the excel sheet.

im-deepfriedwater commented 6 years ago

image

Alrighty, so all those cells with the green in the corner seem to indicate text inputted as opposed to numbers which is causing the NaNs to be read-in. Will try to reformat the input sheet and hopefully that fixes it!

im-deepfriedwater commented 6 years ago

Okay, so I've corrected all three sheets! They seem to working so far, as GRNmap gets to the point where it starts generating graphs. I am unable to test the an entire run for each (my poor laptop takes quite a bit of time to complete a run). To get 100% confidence I'll have to take GRNmap and run these sheets on a different computer, perhaps the ones in LSB when I get a chance. In the end, it did seem to only be a result of the input sheet getting text inputted into the number cells.

But will try to run the sheets later in the week completely so we can consider this issue resolved!

(corrected)-16-genes_27-edges_db6_Sigmoid_estimation_all-timepoints.xlsx (corrected)-17-genes_32-edges_db3_Sigmoid_estimation_all-timepoints_no-dHAP4-data.xlsx (corrected)-15-genes_34-edges_db7_Sigmoid_estimation_all-timepoints_no-dHAP4-data.xlsx

kdahlquist commented 6 years ago

Aaahhhh. I should have known. I will run the corrected workbooks today and post the results (probably tomorrow).

kdahlquist commented 6 years ago

Two of the three files have completed the runs; the third one is still running. I'm going to close this one.