I was looking that kind of “Really” simple stuff only!!

If anybody could elaborate at all on what “adjusting accordingly” means, I’d really appreciate it.

Hey, I updated the script with your suggestions and it’s almost working – now I just get this error however:
Traceback (most recent call last):
File “GoogleTrendsSlopeCalculator.py”, line 26, in
connector.request_report(payload)
AttributeError: ‘TrendReq’ object has no attribute ‘request_report’

Might you know how to fix this?

and its work .. 🙂

thanks for the script

Hi! Actually you can download csv report without being connected. I made a crappy python script with two functions (get_volume_keyword_year, get_volume_keyword_month). Just choose your keyword and that’s it.
https://pastebin.com/raw/i89DsQXc

Sorry pastebin broke the indentation

Okay, just looked at the documentation. Ignore my previous comment. They changed somethings. You’ll have to adjust accordingly unfortunately.

Maybe try installing from github? https://github.com/GeneralMills/pytrends

I then updated connector to be: connector = TrendReq(google_username, google_password, hl=’en-US’, tz=360, custom_useragent=None)

However, this then leads to problems with connector not having the proper attributes. Any ideas?

You authenticate so it allows you to download the CSV.

For how it’s being used in my process, check out the SlideShare embedded in the post, it explains 🙂

I’ve not pushed it to the point of getting locked out yet. If you hit a limit, I’d love to know. You can use categories by changing the payload variable like (untested):

payload = {'geo': 'US', 'q': [row[0]], 'cat': '0-71'}

It’s not mixing up the browser as far as I’m aware, but the trends downloading is being handled by a 3rd party library which I haven’t dissected.

KNIME is open source and Rapidminer is not, so KNIME is my preference 🙂