The NASEM 2021 Nutrient Requirements of Dairy Cattle Model in Python
This is a python version of the Nutrient Requirments of Dairy Cattle 8th revisied edition (NASEM 2021)[https://nap.nationalacademies.org/catalog/25806/nutrient-requirements-of-dairy-cattle-eighth-revised-edition] model. The equations have been copied from the R functions released with the official NASEM software: https://nap.nationalacademies.org/resource/25806/Installation_Instructions_NASEM_Dairy8.pdf
License
This project is distributed under the terms of the MIT License.
Code Attribution
The NASEM equations were coppied from the R code that was distributed under different licensing terms. The original R code is distributed under a separate license.
Please refer to both the MIT License and the R Code License for detailed terms.
Contributing
We welcome contributions! Please read our Contributing Guidelines to get started.
Installation
This package can be installed through a terminal using:
pip install nasem_dairyThe environment that this is installed into will require
python >= 3.10. It will also install pandas and numpy
automatically.
The suggested way to install this is inside a virtual environment. For example, by using Miniconda.
-
Open a terminal window, e.g., Anaconda Prompt.
-
Create a new conda environment with Python installed:
conda create --name myenvironment python>=3.10 conda activate myenvironment -
Install this package via pip:
pip install nasem_dairy
Using the Model
Import package
First, import nasem_dairy. pandas is also required if you want to modify the user_diet input.
import nasem_dairy as nd
import pandas as pdGet input data
There are several demo scenarios built into the app which can be used as input to the
model. These can be acessed using nd.demo():
user_diet, animal_input, equation_selection, infusion_input = nd.demo("input")These 4 inputs can be modified to simulate different scenarios:
- user_diet (pd.DataFrame): The name and kg/d to feed for each ingredient in the diet.
- animal_input (dict): Parameters to descibe the animal being modelled.
- equation_selection (dict): Select which equation to use when there are multiple options.
- infusion_input (dict): If simulating infusions specify this here. This is optional.
Execute Model
To run the model use nd.nasem and pass the input data. This will return an instance of the nd.ModelOutput.
Printing this will provide a snapshot of the results.
## RUN MODEL ###
output = nd.nasem(
user_diet = user_diet_in,
animal_input = animal_input_in,
equation_selection = equation_selection_in,
infusion_input = infusion_input)
print(output)=====================
Model Output Snapshot
=====================
Milk production kg (Mlk_Prod_comp): 34.197
Milk fat g/g (MlkFat_Milk): 0.053
Milk protein g/g (MlkNP_Milk): 0.037
Milk Production - MP allowable kg (Mlk_Prod_MPalow): 35.906
Milk Production - NE allowable kg (Mlk_Prod_NEalow): 31.261
Animal ME intake Mcal/d (An_MEIn): 59.976
Target ME use Mcal/d (Trg_MEuse): 52.195
Animal MP intake g/d (An_MPIn_g): 2565.214
Animal MP use g/d (An_MPuse_g_Trg): 1989.985
Animal RDP intake g/d (An_RDPIn_g): 3556.058
Diet DCAD meq (An_DCADmeq): 96.453
This is a `ModelOutput` object with methods to access all model outputs. See help(ModelOutput).Exploring the Results
The nd.ModelOutput class provides several methods to help you explore the results.
If you need a specific value you can used get_value().
Du_MiCP = output.get_value("Du_MiCP")
print(Du_MiCP)2.0524052134583415If you don't know the exact name of the variable you can use the search() method.
For example if you want to see values related to microbial crude protein (MiCP) you
could search for "MiCP".
print(output.search("MiCP")) Name Value Category Level 1 Level 2
0 Du_EndCP 0.280988 Production MiCP Du_EndCP
1 Du_EndCP_g 280.98834 Production MiCP Du_EndCP_g
2 Du_EndN 0.04507 Production MiCP Du_EndN
3 Du_EndN_g 45.07041 Production MiCP Du_EndN_g
4 Du_IdEAAMic 767.796581 Production MiCP Du_IdEAAMic
5 Du_MiCP 2.052405 Production MiCP Du_MiCP
6 Du_MiCP_g 2052.405213 Production MiCP Du_MiCP_g
7 Du_MiN_NRC2001_g 334.939405 Production MiCP Du_MiN_NRC2001_g
8 Du_MiN_g 328.384834 Production MiCP Du_MiN_g
9 Du_MiTP 1.691182 Production MiCP Du_MiTP
10 Du_MiTP_g 1691.181896 Production MiCP Du_MiTP_g
11 Du_NANMN_g 302.751473 Production MiCP Du_NANMN_g
12 Du_NAN_g 631.136307 Production MiCP Du_NAN_g
13 Du_idMiCP 1.641924 Production MiCP Du_idMiCP
14 Du_idMiCP_g 1641.924171 Production MiCP Du_idMiCP_g
15 Du_idMiTP 1.352946 Production MiCP Du_idMiTP
16 Du_idMiTP_g 1352.945517 Production MiCP Du_idMiTP_g
17 EndAAProf [4.61, 2.9, 4.09, 7.67, 6.23, 1.26, 3.98, 5.18... Production MiCP EndAAProf
18 Fe_DEMiCPend 2.319218 Excretion fecal Fe_DEMiCPend
19 Fe_RumMiCP 0.410481 Excretion fecal Fe_RumMiCP
20 MiCP Dictionary Production MiCP
21 MiN_Vm 340.791931 Production MiCP MiN_Vm
22 MiTPAAProf [5.47, 2.21, 6.99, 9.23, 9.44, 2.63, 6.3, 6.23... Production MiCP MiTPAAProf
23 RDPIn_MiNmax 2.94252 Production MiCP RDPIn_MiNmax
24 Rum_MiCP_DigCHO 0.326383 Production MiCP Rum_MiCP_DigCHO
25 SI_dcMiCP 80.0 Inputs coeff_dict SI_dcMiCP
26 fMiTP_MiCP 0.824 Inputs coeff_dict fMiTP_MiCPThis will print a table with all the variables related to the search string and their value. The "Category" and "Level" columns show where the values are stored within ModelOutput. In this case "MiCP" is the name of a dictionary containg values related to microbial crude protein production, so all values in this dictionary are displayed even if "MiCP" is not in the variable name.
The nasem_dairy package also includes all of the reports from the R version. These can be
accessed through the get_report() method by specifying the report name.
print(output.get_report("table1_1")) Description Values
0 Animal Type Lactating Cow
1 Animal Breed Holstein
2 Body Weight, kg 624.795
3 Empty BW, kg 512.395561
4 Birth Weight, kg 44.1
5 Mature Weight, kg 700.0
6 Mature Empty BW, kg 574.0
7 Age, d 820.8
8 Condition Score, 1 to 5 3.0
9 Percent First Parity 100.0
10 Days in Milk 100
11 Age At First Calving, d
12 Days Pregnant 46
13 Temperature, C 22.0
14 Use In vitro NDF digest 0
15 Feeding Monensin, 0=No, 1=Yes 0
16 Grazing 0
17 Topography 0.0
18 Dist. (Pasture to Parlor, m) 0.0
19 One-Way Trips to the Parlor, m 0The NASEM Directed Acyclic Graph (DAG)
nasem_dairy comes with an optional dag subpackage. This subpackage requires you install
the graph-tool package. This package has several requirements depending on your operating system.
For instructions on how to install visit their website
This subpackage parses the equations used to build the NASEM model and uses the function arguments to build a DAG. This can then be used to explore how varaibles in the model are connected.
Once you have graph-tool installed you can create a DAG.
dag = nd.ModelDAG()Once you have created the DAG you can explore how functions are connected using dag.get_calculation_order().
dag.get_calculation_order("GrUter_Wt")Requirements for Calculating GrUter_Wt
Order of Functions to Call:
1. calculate_Uter_Wtpart
2. calculate_Uter_Wt
3. calculate_GrUter_Wtpart
4. calculate_GrUter_Wt
Required User Inputs:
animal_input:
- An_AgeDay
- An_GestLength
- An_Parity_rl
- An_GestDay
- Fet_BWbrth
- An_LactDay
Required Constants:
coeff_dict:
- GrUter_Ksyn
- GrUterWt_FetBWbrth
- Uter_KsynDecay
- GrUter_KsynDecay
- Uter_Ksyn
- Uter_Kdeg
- Uter_Wt_coeff
- UterWt_FetBWbrthThis will display the functions that need to be called in order to calulate the specified value. It also provides all of the input values that are involved in the calculation.
If you are only interested in calcualting a specific value you can use the create_function() method.
This will return a function that takes the required inputs and will call all of the functions
required to calculate the specified value. This returns a ModelOuptut just like the nd.nasem() function.
GrUter_Wt_function = dag.create_function("GrUter_Wt")
print(inspect.signature(GrUter_Wt_function))(animal_input, coeff_dict)By inspecting the signature of the created function we see it requires animal_input and coeff_dict.
The animal_input can be the same as the one used to run nd.nasem(), though only the specified keys are required.
The coeff_dict contains default coefficients. The default can be accessed as nd.coeff_dict.