topic Re: Sorting before continuing in Python Questions

Sorting before continuing

TomKearns — Tue, 16 Jul 2013 13:53:43 GMT

I am interested in sorting via an attribute before running my script. Currently I am trying to use an update cursor:

arcpy.gp.UpdateCursor (in_dataset,"","","", "attribute D")

It does not appear to work when I use the next search cursor. I sort as above then call the next line, check the fid and it is still 0. I tried to refresh in an attempt to make the sort stick but as of yet it has not worked.

Do I have to use "Sort_management"? I would like to complete the operation without creating a new .shp. It seems that the lock Sort_management puts on the file prevents it from writing over its self.

Thanks!

Re: Sorting before continuing

MathewCoyle — Tue, 16 Jul 2013 13:57:09 GMT

Are you trying to use a da cursor?

sorted(arcpy.da.SearchCursor(table, 'atribute'))

Re: Sorting before continuing

TomKearns — Tue, 16 Jul 2013 14:03:29 GMT

I was not. When I started the script I was unsure if it would be run on machines with only 10 so I used the old update cursor.

Re: Sorting before continuing

TomKearns — Sun, 12 Dec 2021 02:50:03 GMT

I see the .da has an editor session available. If I use this will I not have to create a second .shp?

Something along the lines of:

 with arcpy.da.Editor(workspace) as edit:
        arcpy.Sort_management (in_dataset, out_dataset, sort_field)

If this is the case, how do I close the edit session?

Re: Sorting before continuing

MathewCoyle — Tue, 16 Jul 2013 14:12:21 GMT

Is 'attribute' the name of your field? You should include more of your code to see what you are trying to accomplish.

Re: Sorting before continuing

TomKearns — Sun, 12 Dec 2021 02:50:06 GMT

The LINKNO attribute counts up to 449 and it is the line with 449 as the attribute that I want the script to read first. When I call the fid it still reads as fid 0, not 115 as I was shooting for.

            arcpy.gp.UpdateCursor (stream_in,"","","", "LINKNO D")
            count1 = arcpy.gp.GetCount_management(stream_in)
            cursor = arcpy.SearchCursor(stream_in)
            row = cursor.next()
            fid = row.FID
            print "naming fid"
            print fid

Re: Sorting before continuing

MathewCoyle — Tue, 16 Jul 2013 14:43:46 GMT

Are you trying to extract one particular row from your data? Or are you trying to start at an arbitrary row and continue from that?

Re: Sorting before continuing

TomKearns — Tue, 16 Jul 2013 14:50:37 GMT

I am trying to start at the highest LINKNO and the script will then move through the table row by row. This script will be run on multiple shape files and there is no way to tell ahead of time which fid the highest LINKNO will correspond to.

Re: Sorting before continuing

RhettZufelt — Sun, 12 Dec 2021 02:50:09 GMT

The LINKNO attribute counts up to 449 and it is the line with 449 as the attribute that I want the script to read first. When I call the fid it still reads as fid 0, not 115 as I was shooting for.
            arcpy.gp.UpdateCursor (stream_in,"","","", "LINKNO D")
            count1 = arcpy.gp.GetCount_management(stream_in)
            cursor = arcpy.SearchCursor(stream_in)
            row = cursor.next()
            fid = row.FID
            print "naming fid"
            print fid

See this post for syntax using the sort in a normal cursor. Examples for both hard coding the attribute and using a variable for it. http://forums.arcgis.com/threads/88186-To-select-a-record-based-on-the-highest-value-of-a-field

also, I see you are sorting your updateCursor, yet you are grabbing the cursor.next() on the cursor object created by the searchcursor. I don't see a sort on the searchcursor, so should get the first fid each time.
Perhaps try setting the cursor = to the update cursor and see if it grabs the correct value.

R_

Re: Sorting before continuing

TomKearns — Tue, 16 Jul 2013 15:27:51 GMT

Thank you, I had just found that same thread a few minutes ago and am trying to grab the FID using it. There is something wrong with my syntax though as it keeps crashing. I understand what you were saying about calling the wrong cursor, and will look at that as well.

maxValue = arcpy.gp.SearchCursor(stream_in,"","","", "LINKNO D").next().getValue(FID)

Re: Sorting before continuing

TomKearns — Sun, 12 Dec 2021 02:50:11 GMT

Ok, here is working code to choose the appropriate fid. Thanks for the help everybody.

            uCursor = arcpy.gp.UpdateCursor (stream_in,"","","", "LINKNO D")
            urow = uCursor.next()
            fidStart = urow.FID
            print "fid start"
            print fidStart

Re: Sorting before continuing

TomKearns — Tue, 16 Jul 2013 15:49:39 GMT

That still didn't answer the question about sorting before processing though. Now I can grab the fid of the highest value but I am still interested in sorting the whole table based on the values in the LINKNO column.

Re: Sorting before continuing

RhettZufelt — Sun, 12 Dec 2021 02:50:14 GMT

            uCursor = arcpy.gp.UpdateCursor (stream_in,"","","", "LINKNO D")
            urow = uCursor.next()
            fidStart = urow.FID
            print "fid start"
            print fidStart

Not sure what you mean here. They way you have it above, the uCursor object IS your entire table sorted by LINKNO descending. the uCursor.next() just grabs the next (in this case the first) value.

for row in uCursor:
    print row.FID

Should print all your FID's sorted descending.

R_

Re: Sorting before continuing

TomKearns — Tue, 16 Jul 2013 16:39:51 GMT

Yes but my origional goal was to sort the whole table based on the LINKNO attribute and then run the rest of the script with the table sorted as such (the script creates a new attribute and fills it according to other variables). That is why my initial cursor was in an UpdateCursor as opposed to SearchCursor. What I was finding was that the table was no longer sorted when I proceded with the script. In thinking about it now, it makes sense due to it being only the cursor that is sorted, not the table.

To restate my problem hopefully to improve clarity:

I am interested in sorting my table by an attribute and then proceding with my script. I was attempting to do it using an UpdateCursor but found when I called the FIDs that the table was not sorted. Ideally I could accomplish the sorting without creating a new .shp.

Re: Sorting before continuing

RhettZufelt — Tue, 16 Jul 2013 16:57:57 GMT

Your sorted updateCursor will allow you to "update" values in a field, and would be based on the "sorted" table.

I.e, if you setValue of a new field = 1000, then when you look at your table, after the update, the "last" record (highest FID) would be updated to 1000 (since you've done a cursor.next(), you have the "first" row (in decending order) selected so would actually be the "last" row of the table).

So, if you create the new attribute field before you establish the updateCursor, you can easily update the record based on the sorted table. Just do your updates before you exit the sorted updateCursor.

however, you can't make a table sort of a feature class permenant without creating a new dataset.

The "order" of a shapefile is determined by the order that the features are added. If you are actually after a "sorted" shapefile, not just a sorted table, you would want to use the arcpy.sort function which does create a new feature class.

So, could sort the shapefile into a new fc, delete your existing shapefile, then convert the FC back to shapefile with your original name. That is, of course, if you really, for some reason need a shapefile, otherwise, you can just use the new FC (since output of Sort doesn't support shapefiles).

http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//001700000057000000
http://resources.arcgis.com/en/help/main/10.1/index.html#//001700000057000000

R_

If this doesn't get you going, maybe some input of what the other variables would be, and how you want it "filled" would give me a better idea. Are these variables coming from other fields, static input, etc.?

Re: Sorting before continuing

TomKearns — Tue, 16 Jul 2013 17:22:43 GMT

My problem may have been that I was exiting the update cursor and expecting the order to stick. I am not sure how to stay within the cursor, I have bunch of for loops that run after the reordering.

The rest of the code looks at the variable I am sorting by and another and then assigns a name to the line. It then queries the next row and if it is part of the link it assigns it the same name, all the way down to 0. Then it starts over at then next link without a name and repeats. The cause for the question is that the longest line of links needed to be named first, they always have the highest LINKNO but do not necessaraly fall at FID 0.

LINKNO UPLINK NAME
449         448      a
448         446      a
447         445      b
446         410      a
445         403      b         and so on.

I gave in and just created a new .shp with "arcpy.Sort_management", it adds hastle but once coded it will be fine.