furter | building page-object like structures | iOS library

The problem was in right flags for CGO. This is how header for CGO in main.go looks now

I suspect this issue is that you are expecting this function to allocate grph for you then operating on the allocated graph from the calling code (you dont show the calling code)

ie you are doing something like

Could you please try following, written and tested with shown samples in GNU awk.

This provides the correlation matrix unravelled into a row for each date. There will be n*n columns where ts has n columns:

I think a little explanation about how Android runs your code might help.

So in the background, Android is running some code in an endless loop. Part of that loop is to check a message queue, which is basically where tasks get delivered - code it needs to execute. Ideally it will get all the work required in each loop finished quickly enough that it can manage 60 loops per second. If it has too much to do, that's where it starts to slow down and feel janky.

Some of the tasks that might get executed are things like an Activity being created, so it might want the system to run its onCreate code - which happens once for each Activity. You've probably noticed it only happens once, and that's where you've put your endless loop, right? To kind of trap the execution in there, until the AsyncTask delivers its result?

The problem is you're stopping the main loop from working - it can't move on and do anything until it's finished that task, and you're blocking it on purpose. That's very bad in general, and it's why you're encouraged not to do slow operations on that main thread (the one that runs the main looper) - it creates work that takes too long, and makes the whole UI less responsive (UI handling is also tasks that need to run)

So that's bad in general, but the way AsyncTasks actually work is they run on another thread (so they're not blocking the main one, it's like having another independent person working on stuff) - but they deliver the result on the main thread. They post a message to the main thread's message queue, telling it to run the onPostExecute code.

But the system can't get to that message until it's handled the earlier tasks in the queue. And if it's busy running an endless loop waiting for the result of the AsyncTask, it will never get to that message, the variable will never be updated, and the loop will never see the change it's waiting for. It's like someone holding up a line, refusing to move until they see something that someone further down the line is trying to deliver

That's a bunch of background, but the point is you shouldn't ever block a thread like that, unless it's a special thread you created so you know it's ok and it's not interfering with anything else. Think of Android more as an event-driven system - you write code that should be executed when something happens, like when an Activity gets created (ònCreate) or when a button is pressed(onClick) or when an AsyncTask completes (onPostExecute). There's a reason all these methods start with "on"! "When a thing happens, do this..."

So when your task completes, it runs onPostExecute and that's where all your code to handle receiving the result should go. It doesn't literally need to be inside the onPostExecute method - like you've put yours inside a passJSONGet method to keep things organised, but that gets called from onPostExecute, that's the event that triggers the code being run.

So whatever you need to do with the result, call it from onPostExecute. When that happens, it will do the stuff you've told it to do. Update a variable, make a toast, populate views in your layout with some new data, whatever!

Probably your dataframe doesn't contain the information you're expecting. Or maybe the F4 part has a completely different range? If e.g. the values of Km, Liquid are 10 times smaller, they would get reduced to just a line near zero.

You could print out df2['Counts'][471:494] and df2['Counts'][495:] to see how the values look like. The code below adds a scatter plot to see all the points.

Some remarks:

• Slicing as in array[a:b] goes from a till b-1 (similar to python's range function). So you might want to use data = [df2['Counts'][0:471], df2['Counts'][471:495],df2['Counts'][495:]] or so.
• It is recommended not to mix the ax. and plt. way of calling matplotlib functions.
• Usually it is better to call the xticks functions after creating the main plot, as many functions try to set their own ticks.
• When setting the xticklabels, it is recommended to set the corresponding xticks at the same time. In this case, the ticks are numbered internally as 1, 2, 3.

Here is some example code showing how things could fit together. To help debugging, a scatter plot is added to see where all the points really are.