plot.assess errors and corrections made for DID and ITS models

[szuniga07]

Hello,

I'm sorry to let you know that I found some errors in the plot.assess function. I have corrected the errors and the current developers release I have for ham has the corrections.

I have also allowed general plot options and made graphing parameter options available. And options for arrows and variable names to show how the output coefficients are represented in the different parts of the graph (e.g., the names for the intercept, slopes, etc. can be added to the graphs with arrows pointing in the flow of the trends). Please check out the bottom of, https://github.com/szuniga07/ham, to see examples.

I'm sorry again about this, please download the version on github, I will submit these revisions to CRAN once I make my latest version after completing additional desired features.

Here are the errors (if you're not using the graphs for multiple group Interrupted Time Series, you may not notice the difference).

plot.assess errors:

1. DID=“many” time 1 is no longer the intercept and
2. ITS=“two” intercept is now correct (most noticeable error)

Error of not specifying the correct data frame name in an Interrupted Time Series single-group example:

1. Dataset correctly changed to data=hosp1

interpret() Error for DID "Many" (did="many")

1. Removes the cut-and-paste error in the description (e.g., "ATET") taken from the did="two" model. The DID predictor now correctly reads in the interpretation

[szuniga07]

Corrections have been made. Please use the current developers release found here for revised ham package code:
https://github.com/szuniga07/ham