{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Tutorial\n",
"### Follow along\n",
"\n",
"Code for all the examples is located in your `PYTHONPATH/Lib/site-packages/eonr/examples` folder. With that said, you should be able to make use of `EONR` by following and executing the commands in this tutorial using either the sample data provided or substituting in your own data.\n",
"\n",
"*You will find the following code included in the* `tutorial.py` *or* `tutorial.ipynb` *(for* [Jupyter notebooks](https://jupyter.org/)*) files in your* `PYTHONPATH/Lib/site-packages/eonr/examples` *folder - feel free to load that into your Python IDE to follow along.*"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - -\n",
"### Load modules\n",
"After [installation](installation.md), load `Pandas` and the `EONR` module in a Python interpreter:"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"EONR version: 0.2.0\n"
]
}
],
"source": [
"import os\n",
"import pandas as pd\n",
"import eonr\n",
"print('EONR version: {0}'.format(eonr.__version__))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - -\n",
"### Load the data\n",
"`EONR` uses Pandas dataframes to access and manipulate the experimental data."
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
year
\n",
"
location
\n",
"
plot
\n",
"
trt
\n",
"
rep
\n",
"
time_n
\n",
"
rate_n_applied_kgha
\n",
"
yld_grain_dry_kgha
\n",
"
nup_total_kgha
\n",
"
soil_plus_fert_n_kgha
\n",
"
\n",
" \n",
" \n",
"
\n",
"
0
\n",
"
2012
\n",
"
Minnesota
\n",
"
101
\n",
"
8
\n",
"
1
\n",
"
Pre
\n",
"
235.3785
\n",
"
12410.916200
\n",
"
198.759898
\n",
"
284.69590
\n",
"
\n",
"
\n",
"
1
\n",
"
2012
\n",
"
Minnesota
\n",
"
102
\n",
"
3
\n",
"
1
\n",
"
Pre
\n",
"
67.2510
\n",
"
10627.946000
\n",
"
147.971755
\n",
"
116.56840
\n",
"
\n",
"
\n",
"
2
\n",
"
2012
\n",
"
Minnesota
\n",
"
103
\n",
"
1
\n",
"
1
\n",
"
Pre
\n",
"
0.0000
\n",
"
7428.081218
\n",
"
98.769392
\n",
"
38.10890
\n",
"
\n",
"
\n",
"
3
\n",
"
2012
\n",
"
Minnesota
\n",
"
104
\n",
"
2
\n",
"
1
\n",
"
Pre
\n",
"
33.6255
\n",
"
9202.953180
\n",
"
111.440210
\n",
"
71.73440
\n",
"
\n",
"
\n",
"
4
\n",
"
2012
\n",
"
Minnesota
\n",
"
105
\n",
"
4
\n",
"
2
\n",
"
Pre
\n",
"
100.8765
\n",
"
10841.127180
\n",
"
142.663887
\n",
"
154.67730
\n",
"
\n",
"
\n",
"
5
\n",
"
2012
\n",
"
Minnesota
\n",
"
106
\n",
"
7
\n",
"
2
\n",
"
Pre
\n",
"
201.7530
\n",
"
10646.649330
\n",
"
178.802092
\n",
"
255.55380
\n",
"
\n",
"
\n",
"
6
\n",
"
2012
\n",
"
Minnesota
\n",
"
107
\n",
"
6
\n",
"
2
\n",
"
Pre
\n",
"
168.1275
\n",
"
12367.436000
\n",
"
186.053531
\n",
"
201.75300
\n",
"
\n",
"
\n",
"
7
\n",
"
2012
\n",
"
Minnesota
\n",
"
108
\n",
"
5
\n",
"
2
\n",
"
Pre
\n",
"
134.5020
\n",
"
13366.361700
\n",
"
196.737290
\n",
"
168.12750
\n",
"
\n",
"
\n",
"
8
\n",
"
2012
\n",
"
Minnesota
\n",
"
201
\n",
"
7
\n",
"
1
\n",
"
Pre
\n",
"
201.7530
\n",
"
14232.053480
\n",
"
228.775204
\n",
"
251.07040
\n",
"
\n",
"
\n",
"
9
\n",
"
2012
\n",
"
Minnesota
\n",
"
202
\n",
"
5
\n",
"
1
\n",
"
Pre
\n",
"
134.5020
\n",
"
14384.824980
\n",
"
226.006218
\n",
"
183.81940
\n",
"
\n",
"
\n",
"
10
\n",
"
2012
\n",
"
Minnesota
\n",
"
203
\n",
"
6
\n",
"
1
\n",
"
Pre
\n",
"
168.1275
\n",
"
13592.219290
\n",
"
182.423028
\n",
"
206.23640
\n",
"
\n",
"
\n",
"
11
\n",
"
2012
\n",
"
Minnesota
\n",
"
204
\n",
"
4
\n",
"
1
\n",
"
Pre
\n",
"
100.8765
\n",
"
14091.078390
\n",
"
187.745096
\n",
"
138.98540
\n",
"
\n",
"
\n",
"
12
\n",
"
2012
\n",
"
Minnesota
\n",
"
205
\n",
"
3
\n",
"
2
\n",
"
Pre
\n",
"
67.2510
\n",
"
10739.981390
\n",
"
133.470950
\n",
"
121.05180
\n",
"
\n",
"
\n",
"
13
\n",
"
2012
\n",
"
Minnesota
\n",
"
206
\n",
"
1
\n",
"
2
\n",
"
Pre
\n",
"
0.0000
\n",
"
8375.090921
\n",
"
109.460245
\n",
"
53.80080
\n",
"
\n",
"
\n",
"
14
\n",
"
2012
\n",
"
Minnesota
\n",
"
207
\n",
"
8
\n",
"
2
\n",
"
Pre
\n",
"
235.3785
\n",
"
13797.485850
\n",
"
195.932161
\n",
"
269.00400
\n",
"
\n",
"
\n",
"
15
\n",
"
2012
\n",
"
Minnesota
\n",
"
208
\n",
"
2
\n",
"
2
\n",
"
Pre
\n",
"
33.6255
\n",
"
9713.487469
\n",
"
113.903035
\n",
"
67.25100
\n",
"
\n",
"
\n",
"
16
\n",
"
2012
\n",
"
Minnesota
\n",
"
301
\n",
"
3
\n",
"
3
\n",
"
Pre
\n",
"
67.2510
\n",
"
12579.012170
\n",
"
180.812783
\n",
"
106.48075
\n",
"
\n",
"
\n",
"
17
\n",
"
2012
\n",
"
Minnesota
\n",
"
302
\n",
"
7
\n",
"
3
\n",
"
Pre
\n",
"
201.7530
\n",
"
13604.571780
\n",
"
208.724988
\n",
"
240.98275
\n",
"
\n",
"
\n",
"
18
\n",
"
2012
\n",
"
Minnesota
\n",
"
303
\n",
"
2
\n",
"
3
\n",
"
Pre
\n",
"
33.6255
\n",
"
10185.959390
\n",
"
121.505528
\n",
"
68.37185
\n",
"
\n",
"
\n",
"
19
\n",
"
2012
\n",
"
Minnesota
\n",
"
304
\n",
"
6
\n",
"
3
\n",
"
Pre
\n",
"
168.1275
\n",
"
14305.321460
\n",
"
204.391319
\n",
"
202.87385
\n",
"
\n",
"
\n",
"
20
\n",
"
2012
\n",
"
Minnesota
\n",
"
305
\n",
"
8
\n",
"
4
\n",
"
Pre
\n",
"
235.3785
\n",
"
13929.592020
\n",
"
186.775288
\n",
"
267.88315
\n",
"
\n",
"
\n",
"
21
\n",
"
2012
\n",
"
Minnesota
\n",
"
306
\n",
"
4
\n",
"
4
\n",
"
Pre
\n",
"
100.8765
\n",
"
10975.799250
\n",
"
147.057081
\n",
"
133.38115
\n",
"
\n",
"
\n",
"
22
\n",
"
2012
\n",
"
Minnesota
\n",
"
307
\n",
"
5
\n",
"
4
\n",
"
Pre
\n",
"
134.5020
\n",
"
11338.070290
\n",
"
148.348790
\n",
"
162.52325
\n",
"
\n",
"
\n",
"
23
\n",
"
2012
\n",
"
Minnesota
\n",
"
308
\n",
"
1
\n",
"
4
\n",
"
Pre
\n",
"
0.0000
\n",
"
5821.373521
\n",
"
68.791363
\n",
"
28.02125
\n",
"
\n",
"
\n",
"
24
\n",
"
2012
\n",
"
Minnesota
\n",
"
401
\n",
"
5
\n",
"
3
\n",
"
Pre
\n",
"
134.5020
\n",
"
13755.002370
\n",
"
198.844611
\n",
"
173.73175
\n",
"
\n",
"
\n",
"
25
\n",
"
2012
\n",
"
Minnesota
\n",
"
402
\n",
"
1
\n",
"
3
\n",
"
Pre
\n",
"
0.0000
\n",
"
9077.628329
\n",
"
122.207140
\n",
"
39.22975
\n",
"
\n",
"
\n",
"
26
\n",
"
2012
\n",
"
Minnesota
\n",
"
403
\n",
"
4
\n",
"
3
\n",
"
Pre
\n",
"
100.8765
\n",
"
13760.323240
\n",
"
181.734857
\n",
"
135.62285
\n",
"
\n",
"
\n",
"
27
\n",
"
2012
\n",
"
Minnesota
\n",
"
404
\n",
"
8
\n",
"
3
\n",
"
Pre
\n",
"
235.3785
\n",
"
14896.886860
\n",
"
227.454749
\n",
"
270.12485
\n",
"
\n",
"
\n",
"
28
\n",
"
2012
\n",
"
Minnesota
\n",
"
405
\n",
"
2
\n",
"
4
\n",
"
Pre
\n",
"
33.6255
\n",
"
10551.466010
\n",
"
144.282839
\n",
"
66.13015
\n",
"
\n",
"
\n",
"
29
\n",
"
2012
\n",
"
Minnesota
\n",
"
406
\n",
"
6
\n",
"
4
\n",
"
Pre
\n",
"
168.1275
\n",
"
15621.406530
\n",
"
227.056199
\n",
"
200.63215
\n",
"
\n",
"
\n",
"
30
\n",
"
2012
\n",
"
Minnesota
\n",
"
407
\n",
"
3
\n",
"
4
\n",
"
Pre
\n",
"
67.2510
\n",
"
10950.450720
\n",
"
146.628851
\n",
"
95.27225
\n",
"
\n",
"
\n",
"
31
\n",
"
2012
\n",
"
Minnesota
\n",
"
408
\n",
"
7
\n",
"
4
\n",
"
Pre
\n",
"
201.7530
\n",
"
13838.222530
\n",
"
210.340625
\n",
"
229.77425
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" year location plot trt rep time_n rate_n_applied_kgha \\\n",
"0 2012 Minnesota 101 8 1 Pre 235.3785 \n",
"1 2012 Minnesota 102 3 1 Pre 67.2510 \n",
"2 2012 Minnesota 103 1 1 Pre 0.0000 \n",
"3 2012 Minnesota 104 2 1 Pre 33.6255 \n",
"4 2012 Minnesota 105 4 2 Pre 100.8765 \n",
"5 2012 Minnesota 106 7 2 Pre 201.7530 \n",
"6 2012 Minnesota 107 6 2 Pre 168.1275 \n",
"7 2012 Minnesota 108 5 2 Pre 134.5020 \n",
"8 2012 Minnesota 201 7 1 Pre 201.7530 \n",
"9 2012 Minnesota 202 5 1 Pre 134.5020 \n",
"10 2012 Minnesota 203 6 1 Pre 168.1275 \n",
"11 2012 Minnesota 204 4 1 Pre 100.8765 \n",
"12 2012 Minnesota 205 3 2 Pre 67.2510 \n",
"13 2012 Minnesota 206 1 2 Pre 0.0000 \n",
"14 2012 Minnesota 207 8 2 Pre 235.3785 \n",
"15 2012 Minnesota 208 2 2 Pre 33.6255 \n",
"16 2012 Minnesota 301 3 3 Pre 67.2510 \n",
"17 2012 Minnesota 302 7 3 Pre 201.7530 \n",
"18 2012 Minnesota 303 2 3 Pre 33.6255 \n",
"19 2012 Minnesota 304 6 3 Pre 168.1275 \n",
"20 2012 Minnesota 305 8 4 Pre 235.3785 \n",
"21 2012 Minnesota 306 4 4 Pre 100.8765 \n",
"22 2012 Minnesota 307 5 4 Pre 134.5020 \n",
"23 2012 Minnesota 308 1 4 Pre 0.0000 \n",
"24 2012 Minnesota 401 5 3 Pre 134.5020 \n",
"25 2012 Minnesota 402 1 3 Pre 0.0000 \n",
"26 2012 Minnesota 403 4 3 Pre 100.8765 \n",
"27 2012 Minnesota 404 8 3 Pre 235.3785 \n",
"28 2012 Minnesota 405 2 4 Pre 33.6255 \n",
"29 2012 Minnesota 406 6 4 Pre 168.1275 \n",
"30 2012 Minnesota 407 3 4 Pre 67.2510 \n",
"31 2012 Minnesota 408 7 4 Pre 201.7530 \n",
"\n",
" yld_grain_dry_kgha nup_total_kgha soil_plus_fert_n_kgha \n",
"0 12410.916200 198.759898 284.69590 \n",
"1 10627.946000 147.971755 116.56840 \n",
"2 7428.081218 98.769392 38.10890 \n",
"3 9202.953180 111.440210 71.73440 \n",
"4 10841.127180 142.663887 154.67730 \n",
"5 10646.649330 178.802092 255.55380 \n",
"6 12367.436000 186.053531 201.75300 \n",
"7 13366.361700 196.737290 168.12750 \n",
"8 14232.053480 228.775204 251.07040 \n",
"9 14384.824980 226.006218 183.81940 \n",
"10 13592.219290 182.423028 206.23640 \n",
"11 14091.078390 187.745096 138.98540 \n",
"12 10739.981390 133.470950 121.05180 \n",
"13 8375.090921 109.460245 53.80080 \n",
"14 13797.485850 195.932161 269.00400 \n",
"15 9713.487469 113.903035 67.25100 \n",
"16 12579.012170 180.812783 106.48075 \n",
"17 13604.571780 208.724988 240.98275 \n",
"18 10185.959390 121.505528 68.37185 \n",
"19 14305.321460 204.391319 202.87385 \n",
"20 13929.592020 186.775288 267.88315 \n",
"21 10975.799250 147.057081 133.38115 \n",
"22 11338.070290 148.348790 162.52325 \n",
"23 5821.373521 68.791363 28.02125 \n",
"24 13755.002370 198.844611 173.73175 \n",
"25 9077.628329 122.207140 39.22975 \n",
"26 13760.323240 181.734857 135.62285 \n",
"27 14896.886860 227.454749 270.12485 \n",
"28 10551.466010 144.282839 66.13015 \n",
"29 15621.406530 227.056199 200.63215 \n",
"30 10950.450720 146.628851 95.27225 \n",
"31 13838.222530 210.340625 229.77425 "
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df_data = pd.read_csv(os.path.join('data', 'minnesota_2012.csv'))\n",
"df_data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - -\n",
"### Set column names *(pre-init)*\n",
"*The table containing the experimental data* **must** *have a minimum of two columns:*\n",
"\n",
"* *Nitrogen fertilizer rate*\n",
"* *Grain yield*\n",
"\n",
"`EONR` accepts custom column names. Just be sure to set them by either using `EONR.set_column_names()` or by passing them to `EONR.calculate_eonr()`. We will declare the names of the these two columns as they exist in the `Pandas` dataframe so they can be passed to `EONR` later:"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"col_n_app = 'rate_n_applied_kgha'\n",
"col_yld = 'yld_grain_dry_kgha'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Each row of data in our dataframe should correspond to a nitrogen rate treatment plot. It is common to have several other columns describing each treatment plot (e.g., year, location, replication, nitrogen timing, etc.). These aren't necessary, but `EONR` will try pull information from \"year\", \"location\", and \"nitrogen timing\" for labeling the plots that are generated (as you'll see towards the end of this tutorial).\n",
"\n",
"- - -\n",
"### Set units\n",
"Although optional, it is good practice to declare units so we don't get confused:"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"unit_currency = '$'\n",
"unit_fert = 'kg'\n",
"unit_grain = 'kg'\n",
"unit_area = 'ha'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"These unit variables are only used for plotting (titles and axes labels), and they are not actually used for any computations. Because the initial parameter guess is critically important for fitting the model(s), there is logic built in that adjusts this intitial guess based on the units provided here.\n",
"\n",
"- - -\n",
"### Set economic conditions\n",
"`EONR` computes the _Economic_ Optimum Nitrogen Rate for any economic scenario that we define. All that is required is to declare the cost of the nitrogen fertilizer (per unit, as defined above) and the price of grain (also per unit). Note that the cost of nitrogen fertilizer can be set to zero, and the _Agronomic_ Optimum Nitrogen Rate will be computed."
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"cost_n_fert = 0.88 # in USD per kg nitrogen\n",
"price_grain = 0.157 # in USD per kg grain"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - -\n",
"### Initialize `EONR`\n",
"At this point, we can initialize an instance of `EONR`.\n",
"\n",
"*Before doing so, we may want to set the base directory.* `EONR.base_dir` *is the default location for saving plots and data processed by* `EONR`*. If* `EONR.base_dir` *is not set, it will be set to be a folder named \"eonr_tutorial\" in the current working directory during the intitialization (to see your current working directory, type* `os.getcwd()`*). If you do not wish to use this as your current working directory, it can be passed to the* `EONR` *instance using the* `base_dir` *keyword.*\n",
"\n",
"For demonstration purposes, we will set `EONR.base_dir` to what would be the default folder if nothing were passed to the `base_dir` keyword --> that is, we will choose a folder named \"eonr_advanced_tutorial\" in the current working directory (`EONR` will create the directory if it does not exist).\n",
"\n",
"And finally, to create an instance of `EONR`, pass the appropriate variables to `EONR()`:"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"base_dir = os.path.join(os.getcwd(), 'eonr_tutorial')\n",
"\n",
"my_eonr = eonr.EONR(cost_n_fert=cost_n_fert,\n",
" price_grain=price_grain,\n",
" col_n_app=col_n_app,\n",
" col_yld=col_yld,\n",
" unit_currency=unit_currency,\n",
" unit_grain=unit_grain,\n",
" unit_fert=unit_fert,\n",
" unit_area=unit_area,\n",
" model=None,\n",
" base_dir=base_dir)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Notice the `model` keywork argument was set to `None` (default='quad_plateau'). This keyword is used to define which model to use to fit the experimental data. If `None`, `EONR` will try all supported models (i.e., as of v. 0.2.0, both 'quad_plateau' and 'quadratic' models are supported), and use the model that fits the data best as determined by the highest $\\text{r}^2$."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - -\n",
"### Calculate the EONR\n",
"With `my_eonr` initialized as an instance of `EONR`, we can now calculate the economic optimum nitrogen rate by calling the `calculate_eonr()` method and passing the dataframe with the loaded data:"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"Computing EONR for Minnesota 2012 Pre\n",
"Cost of N fertilizer: $0.88 per kg\n",
"Price grain: $0.16 per kg\n",
"Fixed costs: $0.00 per ha\n",
"Checking quadratic and quadric-plateau models for best fit..\n",
"Quadratic model r^2: 0.72\n",
"Quadratic-plateau model r^2: 0.73\n",
"Using the quadratic-plateau model..\n",
"Economic optimum N rate (EONR): 162.3 kg per ha [130.5, 207.8] (90.0% confidence)\n",
"Maximum return to N (MRTN): $767.93 per ha\n"
]
}
],
"source": [
"my_eonr.calculate_eonr(df_data)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"It may take a few seconds to run, but it will take much longer if you choose to compute the bootstrap confidence intervals in addition to the profile-likelihood confidence intervals. Please see the [Advanced tutorial](advanced_tutorial.html#Bootstrap-confidence-intervals) describing how to compute the bootstrap CIs (`EONR` does not run the bootstrap CIs by default)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"And that's it! The economic optimum for this dataset and economic scenario was **162 kg nitrogen per ha** (with 90% confidence bounds at **131** and **208 kg per ha**) and resulted in a maximum net return of nearly **$770 per ha**.\n",
"\n",
"Notice the $\\text{r}^2$ for the quadratic-plateau model was greater than that of the quadratic model, so the quadratic-plateau model was used (the model used is recorded in `EONR.df_results` and will be displayed in the plot legend - see below).\n",
"\n",
"- - - \n",
"### Plotting the EONR\n",
"This is great, but of course it'd be useful to see our data and results plotted. Do this by calling the ```plot_eonr()``` module and *(optionally)* passing the minimum/maximum values for each axis:"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"
"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"my_eonr.plot_eonr(x_min=-5, x_max=300, y_min=-100, y_max=1400)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"* The blue points are **experimental data** (yield value in \\\\$ per ha as a function of nitrogen rate)\n",
"* The blue line is the best-fit quadratic-plateau model representing **gross return to nitrogen**\n",
"* The red line is the **cost of nitrogen fertilizer**\n",
"* The green line is the difference between the two and represents the **net return to nitrogen**\n",
"* The green point is the **Economic Optimum Nitrogen Rate (EONR)**\n",
"* The transparent grey box surrounding the EONR/MRTN (green point) illustrates the **90\\% confidence intervals**\n",
"\n",
"The EONR is the point on the x-axis where the net return curve (green) reaches the maximum return. The return to nitrogen at the EONR is the **Maximum Return to Nitrogen (MRTN)**, indicating the profit that is earned at the economic optimum nitrogen rate.\n",
"\n",
"Notice the economic scenario (i.e., grain price, nitrogen fertilizer cost, etc.) and the \"Base zero\" values in the upper right corner describing the assumptions of EONR calculatioon. *Base zero* refers to the initial y-intercept of the gross return model (this setting can be turned on/off by setting `EONR.base_zero` to `True` or `False`. See the [Advanced tutorial](advanced_tutorial.html#Turn-base_zero-off) and the [API](my_eonr.html#module-eonr.eonr) for more information.\n",
"\n",
"- - -\n",
"### Accesing complete results\n",
"All results (e.g., EONR, MRTN, $\\text{r}^2$ and root mean squared errors from best-fit models, confidence intervals, etc.) are stored in the `EONR.df_results` dataframe:"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
price_grain
\n",
"
cost_n_fert
\n",
"
cost_n_social
\n",
"
costs_fixed
\n",
"
price_ratio
\n",
"
unit_price_grain
\n",
"
unit_cost_n
\n",
"
location
\n",
"
year
\n",
"
time_n
\n",
"
...
\n",
"
mrtn
\n",
"
grtn_r2_adj
\n",
"
grtn_rmse
\n",
"
grtn_max_y
\n",
"
grtn_crit_x
\n",
"
grtn_y_int
\n",
"
scn_lin_r2
\n",
"
scn_lin_rmse
\n",
"
scn_exp_r2
\n",
"
scn_exp_rmse
\n",
"
\n",
" \n",
" \n",
"
\n",
"
0
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
0.0
\n",
"
5.605096
\n",
"
$ per ha
\n",
"
$ per kg
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
...
\n",
"
767.930477
\n",
"
0.728926
\n",
"
181.225993
\n",
"
917.43358
\n",
"
177.439985
\n",
"
0.002964
\n",
"
None
\n",
"
None
\n",
"
None
\n",
"
None
\n",
"
\n",
" \n",
"
\n",
"
1 rows × 33 columns
\n",
"
"
],
"text/plain": [
" price_grain cost_n_fert cost_n_social costs_fixed price_ratio \\\n",
"0 0.157 0.88 0.0 0.0 5.605096 \n",
"\n",
" unit_price_grain unit_cost_n location year time_n ... mrtn \\\n",
"0 $ per ha $ per kg Minnesota 2012 Pre ... 767.930477 \n",
"\n",
" grtn_r2_adj grtn_rmse grtn_max_y grtn_crit_x grtn_y_int scn_lin_r2 \\\n",
"0 0.728926 181.225993 917.43358 177.439985 0.002964 None \n",
"\n",
" scn_lin_rmse scn_exp_r2 scn_exp_rmse \n",
"0 None None None \n",
"\n",
"[1 rows x 33 columns]"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"my_eonr.df_results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"See the [Advanced tutorial](advanced_tutorial.html#View-results) for a description of every column."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - -\n",
"### Visualizing all confidence intervals\n",
"By default, the confidence intervals (CIs) are calculated at many alpha levels. Noting that $\\text{CI} = 1-\\alpha$, let's plot the **Wald** CIs and **profile-likelihood** CIs for a range of $\\alpha$ values.\n",
"\n",
"`EONR.plot_modify_size()` can be called to make basic adjustments to the figure size. Because ``EONR.fig_tau.fig`` is a Matplotlib figure object, you'll be able to make any customizations within the Matplotlib API as well. "
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"
"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"my_eonr.plot_tau()\n",
"my_eonr.fig_tau = my_eonr.plot_modify_size(fig=my_eonr.fig_tau.fig, plotsize_x=5, plotsize_y=4.0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This plot shows the lower and upper confidence intervals of the *True EONR* (*True EONR* refers to the actual EONR value, which is not actually known due to uncertainty in the dataset). At 0\\% confidence, the *True EONR* is the *maximum likelihood* value, but as we increase the confidence level from 67\\%, 80\\%, 90\\%, 95\\%, and 99\\%, the statistical range of the *True EONR* widens.\n",
"\n",
"In general, the profile-likelihood CIs are considered the most accurate of the three methods because they reflect the actual, often asymmetric, uncertainty in a parameter estimate [Cook & Weisberg, 1990](https://www.tandfonline.com/doi/abs/10.1080/01621459.1990.10476233).\n",
"\n",
"- - -\n",
"### Accessing complete CI results\n",
"All data relating to the calculation of the CIs are saved in the `EONR.df_ci` dataframe:"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
run_n
\n",
"
location
\n",
"
year
\n",
"
time_n
\n",
"
price_grain
\n",
"
cost_n_fert
\n",
"
cost_n_social
\n",
"
price_ratio
\n",
"
f_stat
\n",
"
t_stat
\n",
"
level
\n",
"
wald_l
\n",
"
wald_u
\n",
"
pl_l
\n",
"
pl_u
\n",
"
boot_l
\n",
"
boot_u
\n",
"
opt_method_l
\n",
"
opt_method_u
\n",
"
\n",
" \n",
" \n",
"
\n",
"
0
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
0.000000
\n",
"
0.000000
\n",
"
0.000
\n",
"
NaN
\n",
"
NaN
\n",
"
NaN
\n",
"
NaN
\n",
"
NaN
\n",
"
NaN
\n",
"
N/A
\n",
"
N/A
\n",
"
\n",
"
\n",
"
1
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
0.016070
\n",
"
0.126767
\n",
"
0.100
\n",
"
158.247550
\n",
"
166.432039
\n",
"
159.432270
\n",
"
165.322890
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
2
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
0.065374
\n",
"
0.255684
\n",
"
0.200
\n",
"
154.085886
\n",
"
170.593703
\n",
"
156.557488
\n",
"
168.432973
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
3
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
0.151446
\n",
"
0.389161
\n",
"
0.300
\n",
"
149.777004
\n",
"
174.902584
\n",
"
153.670496
\n",
"
171.730580
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
4
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
0.281127
\n",
"
0.530214
\n",
"
0.400
\n",
"
145.223559
\n",
"
179.456029
\n",
"
150.721517
\n",
"
175.297880
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
5
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
0.466549
\n",
"
0.683044
\n",
"
0.500
\n",
"
140.289960
\n",
"
184.389628
\n",
"
147.647498
\n",
"
179.255088
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
6
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
0.729644
\n",
"
0.854192
\n",
"
0.600
\n",
"
134.765004
\n",
"
189.914585
\n",
"
144.357366
\n",
"
183.796319
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
7
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
0.969292
\n",
"
0.984526
\n",
"
0.667
\n",
"
130.557584
\n",
"
194.122004
\n",
"
141.960754
\n",
"
187.330059
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
8
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
1.113663
\n",
"
1.055302
\n",
"
0.700
\n",
"
128.272818
\n",
"
196.406771
\n",
"
140.698855
\n",
"
189.275802
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
9
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
1.719858
\n",
"
1.311434
\n",
"
0.800
\n",
"
120.004454
\n",
"
204.675134
\n",
"
136.361782
\n",
"
196.472716
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
10
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
2.887033
\n",
"
1.699127
\n",
"
0.900
\n",
"
107.489043
\n",
"
217.190545
\n",
"
130.457577
\n",
"
207.827070
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
11
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
4.182964
\n",
"
2.045230
\n",
"
0.950
\n",
"
96.316254
\n",
"
228.363335
\n",
"
125.804436
\n",
"
218.441868
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
"
\n",
"
12
\n",
"
1
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
5.605096
\n",
"
7.597663
\n",
"
2.756386
\n",
"
0.990
\n",
"
73.358903
\n",
"
251.320685
\n",
"
117.769597
\n",
"
241.793290
\n",
"
NaN
\n",
"
NaN
\n",
"
Nelder-Mead
\n",
"
Nelder-Mead
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" run_n location year time_n price_grain cost_n_fert cost_n_social \\\n",
"0 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"1 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"2 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"3 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"4 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"5 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"6 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"7 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"8 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"9 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"10 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"11 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"12 1 Minnesota 2012 Pre 0.157 0.88 0.0 \n",
"\n",
" price_ratio f_stat t_stat level wald_l wald_u \\\n",
"0 5.605096 0.000000 0.000000 0.000 NaN NaN \n",
"1 5.605096 0.016070 0.126767 0.100 158.247550 166.432039 \n",
"2 5.605096 0.065374 0.255684 0.200 154.085886 170.593703 \n",
"3 5.605096 0.151446 0.389161 0.300 149.777004 174.902584 \n",
"4 5.605096 0.281127 0.530214 0.400 145.223559 179.456029 \n",
"5 5.605096 0.466549 0.683044 0.500 140.289960 184.389628 \n",
"6 5.605096 0.729644 0.854192 0.600 134.765004 189.914585 \n",
"7 5.605096 0.969292 0.984526 0.667 130.557584 194.122004 \n",
"8 5.605096 1.113663 1.055302 0.700 128.272818 196.406771 \n",
"9 5.605096 1.719858 1.311434 0.800 120.004454 204.675134 \n",
"10 5.605096 2.887033 1.699127 0.900 107.489043 217.190545 \n",
"11 5.605096 4.182964 2.045230 0.950 96.316254 228.363335 \n",
"12 5.605096 7.597663 2.756386 0.990 73.358903 251.320685 \n",
"\n",
" pl_l pl_u boot_l boot_u opt_method_l opt_method_u \n",
"0 NaN NaN NaN NaN N/A N/A \n",
"1 159.432270 165.322890 NaN NaN Nelder-Mead Nelder-Mead \n",
"2 156.557488 168.432973 NaN NaN Nelder-Mead Nelder-Mead \n",
"3 153.670496 171.730580 NaN NaN Nelder-Mead Nelder-Mead \n",
"4 150.721517 175.297880 NaN NaN Nelder-Mead Nelder-Mead \n",
"5 147.647498 179.255088 NaN NaN Nelder-Mead Nelder-Mead \n",
"6 144.357366 183.796319 NaN NaN Nelder-Mead Nelder-Mead \n",
"7 141.960754 187.330059 NaN NaN Nelder-Mead Nelder-Mead \n",
"8 140.698855 189.275802 NaN NaN Nelder-Mead Nelder-Mead \n",
"9 136.361782 196.472716 NaN NaN Nelder-Mead Nelder-Mead \n",
"10 130.457577 207.827070 NaN NaN Nelder-Mead Nelder-Mead \n",
"11 125.804436 218.441868 NaN NaN Nelder-Mead Nelder-Mead \n",
"12 117.769597 241.793290 NaN NaN Nelder-Mead Nelder-Mead "
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"my_eonr.df_ci"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - - \n",
"### Adjusting the economic scenario\n",
"These results were calculated for a specific economic scenario, but the cost of fertilizer and price of grain can be adjusted to run `EONR` for another economic scenario. Just adjust the economic scenario by passing any of:\n",
"\n",
"* `cost_n_fert`\n",
"* `price_grain`\n",
"* `cost_n_social`\n",
"\n",
"to `EONR.update_econ()`:"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"cost_n_fert = 1.32 # adjusted from $0.88 per kg nitrogen\n",
"my_eonr.update_econ(cost_n_fert=cost_n_fert)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - -\n",
"### Environmental observations\n",
"You'll notice above that we can pass the `cost_n_social` variable to `EONR.update_econ()`. This is becuase `EONR` will calculate the **Socially Optimum Nitrogen Rate (SONR)** if certain environmental data are available. For more information about the **SONR**, refer to the [Background chapter](background.html#The-social-cost-of-nitrogen).\n",
"\n",
"In the same way that `cost_n_fert` was adjusted in the previous code, `cost_n_social` will be set (for the first time):"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
"cost_n_social = 1.10 # in USD per kg nitrogen\n",
"my_eonr.update_econ(cost_n_social=cost_n_social)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- - -\n",
"### Calculate mineralization\n",
"You may have noticed that [the loaded data](tutorial.html#Load-the-data) for this tutorial contains columns for nitrogen uptake (\"nup_total_kgha\") and available nitrogen (\"soil_plus_fert_n_kgha\"). This data can be used to calculate the approximate net mineralization for the zero rate, which can be assumed to consistent across all rates. Total crop available nitrogen can then be assumed to be the sum of fertilizer, soil, and net mineralized nitrogen. We can then calculate the **SONR** as long as the column names are correctly set. \n",
"\n",
"The column names were set for nitrogen fertilizer rate (`col_n_app`) and grain yield (`col_yld`) during the initialization of `EONR`, but they haven't been set for the nitrogen uptake or available nitrogen columns. This can be done (even after initilization of `EONR`) using `EONR.set_column_names()`:"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"def calc_mineralization(df_data, units_fert='kgha'):\n",
" '''\n",
" Calculates mineralization and adds \"crop_available_n\" to df\n",
" '''\n",
" df_trt0 = df_data[df_data['rate_n_applied_kgha']==0].copy()\n",
" df_trt0['mineralize_n'] = (df_trt0['nup_total_kgha'] -\n",
" df_trt0['soil_plus_fert_n_kgha'])\n",
" trt0_mineralize = df_trt0['mineralize_n'].mean()\n",
" crop_n_label = 'crop_n_available_' + units_fert\n",
" df_data[crop_n_label] = df_data['soil_plus_fert_n_kgha'] + trt0_mineralize\n",
" return df_data\n",
"\n",
"df_data = calc_mineralization(df_data)\n",
"\n",
"col_crop_nup = 'nup_total_kgha'\n",
"col_n_avail = 'crop_n_available_kgha'\n",
"\n",
"my_eonr.set_column_names(col_crop_nup=col_crop_nup,\n",
" col_n_avail=col_n_avail)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"`EONR` simply subtracts *end of season total nitrogen uptake* from *available nitrogen* to get **net crop nitrogen use**, which is subsequently used to calculate the **SONR**.\n",
"\n",
"- - -\n",
"### Run `EONR` for the socially optimum rate\n",
"Then simply run `EONR.calculate_eonr()` again to calculate the **SONR** for the updated economic scenario:"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"Computing SONR for Minnesota 2012 Pre\n",
"Cost of N fertilizer: $1.32 per kg\n",
"Price grain: $0.16 per kg\n",
"Fixed costs: $0.00 per ha\n",
"Social cost of N: $1.10 per kg\n",
"Checking quadratic and quadric-plateau models for best fit..\n",
"Quadratic model r^2: 0.72\n",
"Quadratic-plateau model r^2: 0.73\n",
"Using the quadratic-plateau model..\n",
"Socially optimum N rate (SONR): 149.6 kg per ha [122.4, 186.3] (90.0% confidence)\n",
"Maximum return to N (MRTN): $684.60 per ha\n"
]
}
],
"source": [
"my_eonr.calculate_eonr(df_data)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The new results are appended to the `EONR.df_results` dataframe:"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
price_grain
\n",
"
cost_n_fert
\n",
"
cost_n_social
\n",
"
costs_fixed
\n",
"
price_ratio
\n",
"
unit_price_grain
\n",
"
unit_cost_n
\n",
"
location
\n",
"
year
\n",
"
time_n
\n",
"
...
\n",
"
mrtn
\n",
"
grtn_r2_adj
\n",
"
grtn_rmse
\n",
"
grtn_max_y
\n",
"
grtn_crit_x
\n",
"
grtn_y_int
\n",
"
scn_lin_r2
\n",
"
scn_lin_rmse
\n",
"
scn_exp_r2
\n",
"
scn_exp_rmse
\n",
"
\n",
" \n",
" \n",
"
\n",
"
0
\n",
"
0.157
\n",
"
0.88
\n",
"
0.0
\n",
"
0.0
\n",
"
5.605096
\n",
"
$ per ha
\n",
"
$ per kg
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
...
\n",
"
767.930477
\n",
"
0.728926
\n",
"
181.225993
\n",
"
917.43358
\n",
"
177.439985
\n",
"
0.002964
\n",
"
None
\n",
"
None
\n",
"
None
\n",
"
None
\n",
"
\n",
"
\n",
"
1
\n",
"
0.157
\n",
"
1.32
\n",
"
1.1
\n",
"
0.0
\n",
"
15.414013
\n",
"
$ per ha
\n",
"
$ per kg
\n",
"
Minnesota
\n",
"
2012
\n",
"
Pre
\n",
"
...
\n",
"
684.596933
\n",
"
0.728926
\n",
"
181.225993
\n",
"
917.43358
\n",
"
177.439985
\n",
"
0.002964
\n",
"
0.777106
\n",
"
139.314
\n",
"
0.836099
\n",
"
21.1184
\n",
"
\n",
" \n",
"
\n",
"
2 rows × 33 columns
\n",
"
"
],
"text/plain": [
" price_grain cost_n_fert cost_n_social costs_fixed price_ratio \\\n",
"0 0.157 0.88 0.0 0.0 5.605096 \n",
"1 0.157 1.32 1.1 0.0 15.414013 \n",
"\n",
" unit_price_grain unit_cost_n location year time_n ... mrtn \\\n",
"0 $ per ha $ per kg Minnesota 2012 Pre ... 767.930477 \n",
"1 $ per ha $ per kg Minnesota 2012 Pre ... 684.596933 \n",
"\n",
" grtn_r2_adj grtn_rmse grtn_max_y grtn_crit_x grtn_y_int scn_lin_r2 \\\n",
"0 0.728926 181.225993 917.43358 177.439985 0.002964 None \n",
"1 0.728926 181.225993 917.43358 177.439985 0.002964 0.777106 \n",
"\n",
" scn_lin_rmse scn_exp_r2 scn_exp_rmse \n",
"0 None None None \n",
"1 139.314 0.836099 21.1184 \n",
"\n",
"[2 rows x 33 columns]"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"my_eonr.df_results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"`EONR.plot_eonr()` and `EONR.plot_tau()` can be called again to plot the new results:"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"