topic Re: csv.reader to updateCursor in Python Questions

csv.reader to updateCursor

DevinUnderwood2 — Fri, 10 Dec 2021 20:21:06 GMT

How do I modify my syntax to properly read an csv file and write the contents to a shapefile using an update cursor ?

# Read csv file & write to shapefile via update cursor
import csv,os

updateCursor = arcpy.da.UpdateCursor(r'sample.shp', ["Name", "OrderCount","Time","Miles"])

with open(r'CsvToShape.csv','rb') as f:
    reader = csv.reader(f)
    for row in reader:
        route = row[0]
        order = row[1]
        time = row[2]
        mile = row[3]
        for row in updateCursor:
                row.setValue(route = Name)
##                row[1]= order
##                row[2] = time
##                row[3] = mile
                updateCursor.updateRow(row)‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

Re: csv.reader to updateCursor

JoshuaBixby — Tue, 01 May 2018 14:32:57 GMT

You are using the older style cursor, you should look into using What is the data access module?—Help | ArcGIS Desktop .

More importantly, if you want to convert a CSV into a shapefile it is best to have some spatial data. Where is yours or how is it stored in the CSV?

Re: csv.reader to updateCursor

DevinUnderwood2 — Tue, 01 May 2018 14:44:04 GMT

Thanks for the advice.

I did not realize cursors as gone thru the following iterations:

arcpy.UpdateCursor

arcpy.da.UpdateCursor

UpdateCursor

I actually want to read a separate excel csv file and write the data (UpdateCursor) to same named shapefile fields.

Re: csv.reader to updateCursor

VinceAngelo — Tue, 01 May 2018 14:58:27 GMT

You've got a variable name collision on 'row'. The DA cursor row object is just an array/list.

You should probably choose between caching the contents of the CSV in a dictionary and stepping through the UpdateCursor linearly, and using a new cursor with an index-covered WHERE clause for each CSV row. This decision will dictate the structure of your code.

- V

Re: csv.reader to updateCursor

DevinUnderwood2 — Fri, 10 Dec 2021 20:21:09 GMT

Creating a dictionary form the csv is straight forward but I still have the issue of passing the values to the cursor.

import csv,os

updateCursor = arcpy.da.UpdateCursor(r'sample.shp', ["Name", "OrderCount","Time","Miles"])


with open(r'test.csv') as csvfile:
   reader = csv.DictReader(csvfile, delimiter = ",")
   for row in reader:
      updateCursor.updateRow(row)‍‍‍‍‍‍‍‍‍

Re: csv.reader to updateCursor

DanPatterson_Retired — Fri, 10 Dec 2021 20:21:11 GMT

If this is a point file, it is far easier to read the csv file into a numpy array, then use arcpy.NumPyArrayToFeatureClass

Many ways to do this, but a dictionary describing the field names and the data type will do it

dt = {'names': ('Longitude', 'Latitude', 'A', 'B', 'C'),
      'formats': ('f8', 'f8', 'i4', 'f8', 'U4')}

# ---- the csv name, the datatype as described above, skip the header row, delimiter is ','
np.loadtxt('c:/temp/test.csv', dtype=dt, skiprows=1, delimiter=',')

array([(-75.5,  45.5, 1,  1., ' aaa'), (-75. ,  45. , 2,  2., ' bbb')],
      dtype=[('Longitude', '<f8'), ('Latitude', '<f8'), ('A', '<i4'), ('B', '<f8'), ('C', '<U4')])

"""
original file

Longitude, Latitude, A, B, C
-75.5, 45.5, 1, 1.0, aaaa
-75.0, 45.0, 2, 2.0, bbbb

"""‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

Re: csv.reader to updateCursor

RichardFairhurst — Tue, 01 May 2018 23:14:21 GMT

To do a full update there are 3 separate processes. Update records, which occurs as you iterate through the updateCursor records and find a key match in the csv dictionary. Delete records (optional), which occurs as you iterate through the updateCursor records and do not find a key match in the csv dictionary. Update and Delete can be done as you do a single pass on all of the updateCursor records. Finally, Insert records, which I usually approach by first creating a dictionary of the updateCursor records after every record has been updated and deleted. Then I iterate through the csv file or csv dictionary records and if it is not in the updateCursor record dictionary, I insert that records using an InsertCursor into the FC being updated.

If the FC you are updating is empty and all you are doing is adding records, you need to use an insertCursor, not an updateCursor.

Re: csv.reader to updateCursor

JoshuaBixby — Wed, 02 May 2018 14:11:22 GMT

Can you elaborate on "still have the issue of passing the values to the cursor?" Are you getting an error message? If so, what? Are you getting unexpected results? If so, what?

Re: csv.reader to updateCursor

VinceAngelo — Wed, 02 May 2018 14:53:52 GMT

The row object returned by iteration of reader is NOT an array. You must unpack the dictionary into an array in the order required by the field list.

Your current logic assumes that the order of the file and the feature class as processed in the cursor are identical. In most cases, this will end badly.

- V

Re: csv.reader to updateCursor

DevinUnderwood2 — Wed, 02 May 2018 15:03:33 GMT

Thank you for the advice, I will work this out with your suggestions.

Re: csv.reader to updateCursor

DevinUnderwood2 — Wed, 02 May 2018 19:53:10 GMT

It is a polygon.

Re: csv.reader to updateCursor

DevinUnderwood2 — Wed, 02 May 2018 19:58:34 GMT

the polygon feature is not empty, I want to add table info from the csv file.

Re: csv.reader to updateCursor

DevinUnderwood2 — Wed, 02 May 2018 20:19:18 GMT

I believe you are referring to my following example of a dictionary to a list.

In my case, I would read the csv file as a dictionary not hardcoded in the script.

my_dict = {'A1':405,'A2':145}

print str(list_items) + " " + " These are the keys and values"
print str(list_keys) + " " + " These are the keys"
print str(list_values) + " " + " These are values"

[('A1', 405), ('A2', 145)] These are the keys and values
['A1', 'A2'] These are the keys
[405, 145] These are values

Re: csv.reader to updateCursor

VinceAngelo — Fri, 10 Dec 2021 20:21:14 GMT

arcpy DA cursors require that the column list be explicit (or a wildcard, which matches all columns in the order of the table, but it's best practice to always explicitly define your columns), and that the values in the array passed into updateRow must be in the order specified in the cursor declaration. Your code does not do this. It passes in a dictionary, which is forbidden, and generates the error.

In order to get your code working, you'll need to populate an array:

  for row in reader:
      rowArray = []
      for name in ["Name", "OrderCount","Time","Miles"]:
          rowArray.append(row[name])
      updateCursor.updateRow(rowArray)

Be careful, the names might not match by case.

The logic still requires that the rows in the shapefile exactly match the CSV. The safe best is to store the small CSV in a dictionary keyed by ID, add the ID column to the name list (if not keyed by name), and use the shapefile's ID to populate extract the appropriate row.

- V

Re: csv.reader to updateCursor

JumaKhudonazarov — Thu, 31 Jan 2019 18:03:07 GMT

Hi Joshua if I plan to export CSV file and want to see numeric value rather than text how to do that. In other words instead of Yes or Now I want to see 1 or 2?

Re: csv.reader to updateCursor

JoshuaBixby — Thu, 31 Jan 2019 18:19:19 GMT

I suggest starting a new question and providing a bit more information in that new question. For example, what does your data structure look like? I assume the Yes/No column values are currently stored as Text?

Re: csv.reader to updateCursor

JumaKhudonazarov1 — Thu, 31 Jan 2019 18:27:48 GMT

Yes exactly the column values are text should i change it?

Re: csv.reader to updateCursor

JoshuaBixby — Thu, 31 Jan 2019 19:56:07 GMT

Please start a new thread, and I will continue to work with you.

Re: csv.reader to updateCursor

JumaKhudonazarov — Sun, 03 Feb 2019 21:08:27 GMT

Can you please advice can't sorted this out?

Re: csv.reader to updateCursor

JumaKhudonazarov — Fri, 10 Dec 2021 20:21:17 GMT

Hi if I use the can I export CSV file with numeric rather than text? For example instead Yes, or NO. 1 or 2????

# Read csv file & write to shapefile via update cursor
import csv,os

updateCursor = arcpy.da.UpdateCursor(r'sample.shp', ["Name", "OrderCount","Time","Miles"])

with open(r'CsvToShape.csv','rb') as f:
    reader = csv.reader(f)
    for row in reader:
        route = row[0]
        order = row[1]
        time = row[2]
        mile = row[3]
        for row in updateCursor:
                row.setValue(route = Name)
##                row[1]= order
##                row[2] = time
##                row[3] = mile
                updateCursor.updateRow(row