Closed aakankshaduggal closed 3 years ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
View / edit / reply to this conversation on ReviewNB
Shreyanand commented on 2021-04-20T14:49:51Z ----------------------------------------------------------------
I think you accidentally removed the headings... Add the new methods in the content section as well.
aakankshaduggal commented on 2021-04-20T15:52:52Z ----------------------------------------------------------------
Oh my bad. I will add that section with the updated list.
View / edit / reply to this conversation on ReviewNB
Shreyanand commented on 2021-04-20T14:49:51Z ----------------------------------------------------------------
I think if we set the size before plotting it, this will become as big as others as well...
aakankshaduggal commented on 2021-04-20T15:53:07Z ----------------------------------------------------------------
Yeah. I will fix this
View / edit / reply to this conversation on ReviewNB
Shreyanand commented on 2021-04-20T14:49:52Z ----------------------------------------------------------------
The next model we explore is ...
You could also add here that "Linear regression models the linear relationship of predictors whereas ARIMA models the linear relationship of past observation. This makes ARIMA naturally adaptive to new data [source]"
View / edit / reply to this conversation on ReviewNB
Shreyanand commented on 2021-04-20T14:49:53Z ----------------------------------------------------------------
Change this to "This doesn't work. Let's try for orders lesser than 4,4."
aakankshaduggal commented on 2021-04-20T16:29:11Z ----------------------------------------------------------------
Should I just remove this and write a comment instead.
View / edit / reply to this conversation on ReviewNB
Shreyanand commented on 2021-04-20T14:49:54Z ----------------------------------------------------------------
Change to "ARMA(4,3) and ARMA(3,3) have low AIC, with all significant coefficients. Both of these models are good candidates for fitting the series"
View / edit / reply to this conversation on ReviewNB
Shreyanand commented on 2021-04-20T14:49:54Z ----------------------------------------------------------------
Some models are failing to converge. Looking at the logs and the error message, it seems that the optimization algorithm used (LBFGSB) may work differently for different systems. Changing to a simpler model ARIMA(3,1,1) will work here.
@Shreyanand Thanks for the review. The issues with the old models are now resolved.
View / edit / reply to this conversation on ReviewNB
MichaelClifford commented on 2021-04-20T18:45:29Z ----------------------------------------------------------------
do you want to plot this line against the test and training data so readers can see what the output of linear regression looks like? Also if you are going to split it into train and test, do you want to evaluate your linear model on the test data?
aakankshaduggal commented on 2021-04-20T18:47:51Z ----------------------------------------------------------------
We will be covering that bit in the next notebook - forecasting. For now we are just explaining the concepts.
View / edit / reply to this conversation on ReviewNB
MichaelClifford commented on 2021-04-20T18:45:29Z ----------------------------------------------------------------
Can you add plots indicating the time series models generated above?
aakankshaduggal commented on 2021-04-20T18:48:05Z ----------------------------------------------------------------
We will be covering that bit in the next notebook - forecasting. For now we are just explaining the concepts.
View / edit / reply to this conversation on ReviewNB
MichaelClifford commented on 2021-04-20T18:45:30Z ----------------------------------------------------------------
If it doesn't work maybe just remove the commented out cell and this note?
aakankshaduggal commented on 2021-04-20T18:48:19Z ----------------------------------------------------------------
Sure. I will remove this cell.
We will be covering that bit in the next notebook - forecasting. For now we are just explaining the concepts.
View entire conversation on ReviewNB
We will be covering that bit in the next notebook - forecasting. For now we are just explaining the concepts.
View entire conversation on ReviewNB
@MichaelClifford and @Shreyanand I have added the updated notebook with suggested changes :+1:
[APPROVALNOTIFIER] This PR is NOT APPROVED
This pull-request has been approved by: Shreyanand
To complete the pull request process, please assign durandom after the PR has been reviewed.
You can assign the PR to them by writing /assign @durandom
in a comment when ready.
The full list of commands accepted by this bot can be found here.
Related Issues and Dependencies
17
This notebook adds Linear and Exponential models for timeseries data.
This introduces a breaking change
This Pull Request implements
… Explain your changes.
Description