7. Change Log

Changes to each release will be added here

7.1. Release: 0.2.1

Date: 2020 February 19

Description: Added tests and API examples

  • There is now a “tests” folder included in the installation. This can be executed by running python path-to-directory\tests\test.py

  • Examples were added for each EONR class method

  • Added set_column_names, set_units, and set_trial_details methods, so all this information is no longer required to be passed upon initilization of EONR or to the calculate_eonr method.

7.2. Release: 0.2.0

Date: 2019 August 22

Description: Support for a quadratic model

  • If model is not specified, both quadratic and quadratic-plateau models are fit, and the model with the highest r^2 is used

  • The model to be used can be specified using EONR.model (e.g., ‘quadratic’ or ‘quad_plateau’)

  • Removed “bootstrap” text from the “Tau” plot if the bootstrap confidence interval is not computed and there is no bootstrap data to show.

  • Added the model type to EONR.df_results and the legend of the EONR plot.

7.3. Release: 0.1.4

Date: 2019 June 04

Description: Several feature additions and enhancements

  • Consider fixed costs by setting EONR.costs_fixed

  • Added feature to use custom plot title (e.g., EONR.plot_modify_title("My New Title"))

  • Added feature to plot only the NRTN and cost_n_fert line if cost_n_fert is not zero

  • Assert that user has EONR.price_grain > 0 before calculating EONR

  • Added feature to compute the difference from the t-statistic (as a function of theta2/N rate) when calculating profile-likelihood CIs for EONR

  • Added feature to plot the difference from the t-statistic as a function of theta2/N rate (e.g., EONR.plot_delta_tstat)

  • Included “R*” and “costs_at_onr” to EONR.df_results.

  • Taking precaution to be sure optimization of profile-likelihood CIs aren’t caught at a local miniumum.

  • Moving all “linspace” arrays to a new dataframe (EONR.df_linspace)

  • Calculating the derivative of the net return to nitrogen curve (EONR.coefs_nrtn)

  • Added basic plotting for a zoomed in look at the net return curve (EONR.plot_derivative)

  • Added vertical line to represent ONR to EONR.fig_tau

7.4. Release: 0.1.3

Date: 2019 March 31

Description: Fix to profile-likelihood and bootstrap CIs when cost_n_social > 0.

  • Added examples folder to official package distribution

  • Added a fix so EONR.models.R is updated for cost_n_social > 0

  • Column name in df_ci changed from “eonr_error” to “eonr_bias”.

7.5. Release: 0.1.2

Date: 2019 March 30

Description: initial release