I have a data file where a new row is added as the test is processed, not necessarily in chronological order. it also contains data from multiple test locations.
https://imgur.com/a/Tj0Ly2E
Example of my data
I am looking for a way for each value in the series will only be included if the location is the same, so I can have 4 series on the graph each with a handpicked subset of the data.
I made the second graph in the above image by manually rearranging the data, but as this data comes out of a machine I have no way of changing it, and id like to be able to print updated graphs whenever I need to.
in addition I have a multitude of other graphs to and reports to run off this data so I cant manipulate it too much.
something along the logic of,
x series = for(i, count($B$2:$B$13), i++) (if ($A$i = "location_1", include(Sheet1!$B$i)))
I have never seen a function that can do something like this before, am hoping this problem can be solved in some clever way
Related
I have a weird problem that google's not being too kind on results (mostly basic things on how to create a line chart, or combination charts, neither of which I'm after).
I'm trying to automate a report we do each month, I've managed pretty well so far by having a linked excel chart and objects, I have a single sheet in excel I paste my formatted data (set format is consistent each month) then excel picks out the data, creates the charts and when opening powerpoint it reads it all in from there.
Great! Until I was given a new piece of information, there's monthly data and a yearly tally in my data, monthly data is shown with a particular "weighting" applied, year with a different one and it's this that's caused a problem.
The package that's exporting the data can't show these figures contiguously, so we have an export with all the monthly weights, then with annual weights.
So say originally we were looking at C83 to O83 for our data for this year, we're now looking at C83 to N83 and O1364.
I've tried to look at how I would add this to the chart, which currently has the reference:
='Monthly Data to Update'!$C$83:$O$83
But doing anything + or & doesn't seem to work.
So, I decided to just pull all my data in to contiguous set to reference, (so ='Monthly Data to Update'!C83 one cell, ='Monthly Data to Update'!C84 next cell and so on).
This caused a problem for future months that're empty, having 0's in them, so I updated my code to check:
=IF('Monthly Data to Update'!C83=0,NA(),'Monthly Data to Update'!C83)
And this works, my table below my chart is blank for the empty months, my chart doesn't try and plot these empty months. Hurrah!
Now, the wrinkle. The Yearly nets aren't charted but are displayed in the table (I'm not up on Excel charts so how this was originally setup, I just copied the charts from Powerpoint in to Excel and reset the data to where it needed to be before copying it back to Powerpoint as a data linked chart).
First chart I did, works fine, second chart I did it's decided to start plotting the annual net for two of the three lines = \
So, this is how my chart looks like, running off the contiguous (but wrong) data:
And this is how it looks running off the fetched contiguous (correct) data:
It's really baffling, Exceeded and Met now plot the Year to Date figure, whereas Not Met behaves as it should, and they all have the same data layout (on the right) using the same formulas (just pointing at different cells).
I am confused..
I'm hoping to keep this light and simple so have avoided using any macro's to date, so that down the line when I hand this project over to someone any maintenance should be easy to do (say add a chart) without them asking me and me then having to remember what the hell I did a few months ago.. - so if this is doable without going down that line (which, seems like it should be, given my first chart the two lines there worked fine, and one of the three in this is behaving itself).
Apologies for the long winded description, I've given full history so if there's a step further back I could've done better, then happy to fix it up there rather than where I got to.
Ah, so apparently I could've had non-contiguous data with a comma, not a plus or ampersand, so the chart data would be:
=('Monthly Data to Update'!$C$83:$N$83,'Monthly Data to Update'!$O$1404)
So I don't need to rebuild the data in to a contiguous set, and my charts don't cock up when selecting the data this way, win's all around.
Thank you for your time, hopefully this helps someone in the future. (as a saving grace for my lost morning)
I am creating a table using Tabulator, which seems great and very powerful.
I want a way to save relevant data of the table so it can e recreated on the fly.
Currently, I think there are a few things I need...
The row data - I get this using table.getData();
The columns - I get this using table.getColumnDefinitions();
The row data seems perfect I can store that and use it. However, the column information I am saving doesnt appear to have the size of the columns if I have resized them?
Is there a way of getting ALL the relevant column info, so I can save and recreate it exactly?
Alternatively, if there's a single 1 function that saves everything (row data, columns (including order, size etc)) in one go as a JSON or something that may be handy
So you have a few options here.
Config Persistence
If you simply want the table to be the same way it was the last time the user used it on that computer, you could look at using the Peristent Configuration module. This will store a copy of the table column configuration on the browsers local storage so that next time they load the page it will be laid out the same.
Column Layout
If you want to store it externally then you are correct,
the column width is not updated in the definition after a user changes it.
If you want to get the current layout of the columns then you can use the getColumnLayout function to retrieve the current layout of columns:
var columnLayout = table.getColumnLayout();
Though this will only contain the key layout characteristics and not the full definition, you would need to merge them if you wanted to store them in one place.
More details on this method can be found in the Manual Column Layout Documentation
We are trying to track some online marketing metrics and I'm having some trouble. I have 2 tables in different tabs (one imports data from several external data sources, ultimately we want this to be a series of google sheets) and one is the working table.
I have rows on the imported data with month and other attributes defining the data and in the working data these are columns. The working data has a lot of other cells too that are not there with calculations, etc.
What I need to do is to check on the working sheet which month (for example) we are in, then go to the working data and scan all the data for matches with that month. Then I want to consolidate each of the data parameters into the working sheet. Ideally I wouldn't even have to import all the external data into a tab on the working spreadsheet, if I could find a way to work where it would check the external documents for the matches that would be great. The structure of the data in the external documents is the same as displayed here fore the imported data.
Note that in this case it is month but it could be anything random so DATE functions wouldn't work.
So I want to pickup the data from the external source above, and insert in the relevant places. But while the months will not change, other data can change the order in which is imported so we need to check that the headers from picture 2 match the row labels from the imported data.
I hope that makes sense. I would really appreciate any help. Was up until 4 AM trying to figure this out and I would hate to go back to my boss saying he's gonna need to get someone else to do this as I can't. :/
Thank you.
So I resolved this with a INDEX(array, MATCH(),MATCH()) function. First I selected the answer array from the cells with the info I wanted then used the MATCH Function to match the row and the columns I wanted in the matrix.
This created another problem where no answer existed as it threw an error so I had to envelop the whole expression in an IFERROR function.
The final solution was like this:
=IFERROR(INDEX(Table_Query_from_Excel_Files,MATCH(!H1:I1&A1,INDEX(Table_Query_from_Excel_Files[Month]&Table_Query_from_Excel_Files[User location],,),0),MATCH(!A1,Table_Query_from_Excel_Files[#Headers],0)), 0)
I have a existing chart in pptx file, i just want to update its values on daily basis, these charts are already plotted from excel and paste into pptx, i just want to modify the charts ,
but i don't know how many properties are there in charts like series, category and title properties.
i am able to print the chart series and category data, but when i modify these with chart.replace_data(chart_data)
i am getting error like below
return self._chart_part.related_parts[xlsx_part_rId]
KeyError: 'rId4'
I ma using below code to print the values ,
for series in chart.plots:
for i in series.categories:
cat_list.append(i)
for series in chart.series:
for value in series.values:
ser_list.append(value)
the graph looks like below
How can i modify such graphs with updated values which i will be passing from excel to pptx.
I suspect the problem is rooted in the "pasting from Excel" step.
One thing you can try is to right-click on the chart and choose "Edit values in Excel". If that doesn't work for any reason, that would explain the error you're seeing. If it does work, try saving and run the python-pptx code against the saved version.
An alternative, perhaps more reliable fix would be to recreate the chart natively in PowerPoint, starting from adding a new column-chart and adjusting it until you get what you see here. Then you should be able to update it daily with pretty much the code you mentioned (although your ser_list variable will accumulate values for all series in the chart if there is more than one).
To understand why this could be happening, it's important to observe that a chart does not have to have an Excel worksheet "behind" it. It can, and perhaps does in this pasting case, simply have fixed literal values that are embedded in the chart XML itself. The up-side of this is the pasting operation doesn't need to create a whole Excel spreadsheet, which might be tricky in a cut-and-paste context. The down-side is the values can't be conveniently changed using the standard methods.
I haven't tried working with one of those lately so I don't know if editing and saving the data adds a new Excel object created from the embedded values or not, but this might explain the behavior you're seeing.
So I'm using structured references for user data. Eventually we'll have a proper import procedure, but I was hoping to push that back to later and instead use copy-past of their raw data for now.
So I have a table, say tData, that contains a a copy of raw data output. That output already comes with headers, which are actually used in the structured references to the data throughout the workbook.
My problem is that the raw output isn't 100% stable - e.g. some columns may appear & go. Those aren't used in calculations, but it does affect slightly the table structure (position of columns). I can't control the raw output.
I was hoping to instruct them for now to copy-past directly. As I am working with the headers & not position, I thought the formulas would still work.
Summarize in the images below: =tTest[Column2] is the formula set for the box on the right. It should refer to the content of column2 (so 2). If I copy-past different headers, with "Column0" that shifts everything to the right, you can see that a positional reference is actually used by excel to refer to your data. It now returns "1" and Excel even changed the formula to "=tTest[Column1].
Seems wrong to me - e.g. if you reference something by name you don't expect it to actually be referenced by position.
I already tried tTest[[header]:[header]] and it doesn work either.
So, in the end, better write a proper import procedure I think. I won't post here because it's somewhat involved and there are various checks to perform. But overall steps are:
File picker to select the data to input
Copy everything into ThisWorkbook, close input file
Match the column headers from input data with mytable.HeaderRowRange
Copy the raw data into the appropriate column if match if found
So basically write a script to do a match on column headers, which is was I thought Excel would be doing. And still think it should do. However thinking about it, it is probably a lot simpler to code it this way than to have it actually dynamically adapt named references and not positions....
As mentioned by Bob Phillips, you can use INDIRECT. However that does involve adding an extra INDIRECT(reference_to_table_here) to, well, all references. Excel worksheet formulas are clunky enough I think without have to make them any clunkier.
Even though you write the formulas using column names, in the background Excel actually uses position of the column within your table for its calculations. Whenever you change the layout of the workbook, it updates the position for you.