March 21, 2020 at 6:33 pm #183250
Samarendra Chandan Bindu DashParticipant
Not being able to provide correct or up to data is understandable, but the dataset is having errors that defies commonsense. Like take US for example. On 17th March they had 17 people recovered. But on 18th and 19th that number became 0? How ?
This kind of errors are making it harder to fit a model and even score well. Think about it. Say you did all your data cleaning and removed all these errors and then predicted that as on the last day 17 guys had recovered today the recovered number will be at least 17. But no! It is 0.
Handlers please have a look on this case and manage the datasets properly. Thank you.March 21, 2020 at 6:36 pm #183273
Amal V NairKeymaster
Please note the data is not maintained by us and it rightfully belongs to the sources mentioned. We are doing our best to keep the data in sync and remove any discrepancies.
kindly forward fill using pandas to fix such issues.
As the data provided are updated automatically and there can be multiple updates on a single day. The best strategy would be to make use of your 3 submissions.
- Registered Users
- Topic Tags