Tutorial¶

Here we demonstrate the petropandas package. It could be activated by importing pandas alias pd from petropandas. We also import mineral database mindb

In [1]:

from petropandas import pd, mindb

from petropandas import pd, mindb

Matplotlib is building the font cache; this may take a moment.

petropandas are designed as pandas.DataFrame accessors, so we need to create pandas.DataFrame before we can use it. There are several accessors available:

Accessors	Description
petro	petropandas utilities
petroplots	petrological plots
oxides	common calculations with oxides
ree	for rare earth element analyses
isoplot	format data for IsoplotR online
ions	calculations with ions
elements	calculations with elements

Here we're using data from built-in examples, but you can read data from files, clipboard etc. For more details, check https://pandas.pydata.org/docs/reference/io.html

In [2]:

from petropandas.data import bulk as rock
from petropandas.data import minerals as df
from petropandas.data import grt_profile as pro

from petropandas.data import bulk as rock from petropandas.data import minerals as df from petropandas.data import grt_profile as pro

Mineral recalculations¶

Let's see our dataframe...

In [3]:

df

df

Out[3]:

	SiO2	Al2O3	MgO	FeO	MnO	CaO	K2O	Na2O	TiO2	Cr2O3	ZnO	P2O5	Y2O3	F	Cl	Total	Comment
0	37.218	20.349	13.784	13.140	0.000	0.013	8.998	0.404	1.010	0.022	0.020	0.005	0.007	0.367	0.368	95.467	bt-01
1	37.363	20.037	14.106	12.539	0.000	0.013	8.946	0.461	1.206	0.020	0.042	0.000	0.000	0.313	0.335	95.173	bt-02
2	23.748	22.152	9.611	31.549	0.066	0.028	0.023	0.035	0.047	0.000	0.017	0.037	0.000	0.017	0.196	87.475	chl-04
3	46.986	39.183	0.151	1.264	0.033	0.410	3.409	4.883	0.083	0.000	0.000	0.006	0.000	0.000	0.065	96.458	pa-05
4	48.389	32.414	8.227	7.713	0.060	0.027	0.000	0.924	0.005	0.009	0.000	0.005	0.000	0.000	0.014	97.784	cd-06
5	48.870	32.696	8.390	7.482	0.068	0.049	0.000	0.845	0.000	0.000	0.000	0.000	0.000	0.000	0.023	98.418	cd-07
6	61.839	24.661	0.000	0.484	0.015	5.738	0.032	7.917	0.000	0.010	0.019	0.025	0.000	0.000	0.000	100.740	pl-08
7	37.816	21.184	4.189	35.118	1.001	1.435	0.000	0.000	0.000	0.000	0.000	0.006	0.000	0.000	0.004	100.752	g-09
8	37.584	21.311	4.651	33.895	0.999	1.454	0.000	0.035	0.013	0.034	0.000	0.030	0.107	0.000	0.014	100.124	g-10
9	60.657	25.087	0.001	0.000	0.000	5.849	0.106	7.940	0.000	0.000	0.003	0.113	0.000	0.000	0.000	99.756	pl-22
10	61.338	25.025	0.015	0.030	0.003	5.801	0.079	7.997	0.047	0.000	0.057	0.064	0.000	0.000	0.000	100.456	pl-23
11	28.057	53.950	2.040	11.888	0.186	0.036	0.000	0.000	0.585	0.063	1.024	0.000	0.000	0.000	0.007	97.834	st-33
12	27.565	53.885	1.972	11.828	0.175	0.000	0.000	0.000	0.607	0.003	1.035	0.000	0.000	0.000	0.000	97.070	st-34
13	27.091	53.841	2.099	11.875	0.180	0.031	0.000	0.000	0.554	0.046	1.116	0.000	0.000	0.000	0.014	96.844	st-35
14	46.078	35.793	0.652	0.784	0.000	0.002	9.333	1.182	0.697	0.081	0.000	0.000	0.000	0.000	0.013	94.612	ms-49
15	45.487	35.669	0.803	0.895	0.015	0.000	9.394	1.171	0.643	0.033	0.015	0.000	0.034	0.000	0.000	94.159	ms-50

The accessor petro provides several common utilities. Here, we use the search to quickly select data of interest. Of course, any other standard pandas techniques could be used as well.

Let's select analyses of garnet identified by "g-" in column "Comments".

In [4]:

g = df.petro.search("g-", on="Comment")
g

g = df.petro.search("g-", on="Comment") g

Out[4]:

	SiO2	Al2O3	MgO	FeO	MnO	CaO	K2O	Na2O	TiO2	Cr2O3	ZnO	P2O5	Y2O3	F	Cl	Total	Comment
7	37.816	21.184	4.189	35.118	1.001	1.435	0.0	0.000	0.000	0.000	0.0	0.006	0.000	0.0	0.004	100.752	g-09
8	37.584	21.311	4.651	33.895	0.999	1.454	0.0	0.035	0.013	0.034	0.0	0.030	0.107	0.0	0.014	100.124	g-10

wt% could be converted to molar proportions using molprop method:

In [5]:

g.oxides.molprop()

g.oxides.molprop()

Out[5]:

	SiO2	Al2O3	MgO	FeO	MnO	CaO	K2O	Na2O	TiO2	Cr2O3	ZnO	P2O5	Y2O3
7	0.629396	0.207768	0.103935	0.488809	0.014111	0.025590	0.0	0.000000	0.000000	0.000000	0.0	0.000042	0.000000
8	0.625535	0.209013	0.115398	0.471786	0.014083	0.025929	0.0	0.000565	0.000163	0.000224	0.0	0.000211	0.000474

Analyses could be converted to cations p.f.u, either providing number of oxygens:

In [6]:

g.oxides.cations(noxy=12)

g.oxides.cations(noxy=12)

Out[6]:

	Si{4+}	Al{3+}	Mg{2+}	Fe{2+}	Mn{2+}	Ca{2+}	K{+}	Na{+}	Ti{4+}	Cr{3+}	Zn{2+}	P{5+}	Y{3+}
7	3.003379	1.982869	0.495962	2.332521	0.067336	0.122111	0.0	0.000000	0.000000	0.00000	0.0	0.000403	0.000000
8	2.991386	1.999055	0.551848	2.256141	0.067346	0.123994	0.0	0.005401	0.000778	0.00214	0.0	0.002021	0.004532

or number of cations:

In [7]:

g.oxides.cations(ncat=8, tocat=True)

g.oxides.cations(ncat=8, tocat=True)

Out[7]:

	Si{4+}	Al{3+}	Mg{2+}	Fe{2+}	Mn{2+}	Ca{2+}	K{+}	Na{+}	Ti{4+}	Cr{3+}	Zn{2+}	P{5+}	Y{3+}
7	3.001661	1.981734	0.495678	2.331186	0.067297	0.122041	0.0	0.000000	0.000000	0.000000	0.0	0.000403	0.000000
8	2.989651	1.997896	0.551528	2.254832	0.067307	0.123922	0.0	0.005398	0.000778	0.002138	0.0	0.002020	0.004529

or using mineral instance mindb.Garnet from database. Note keeword argument keep to keep extra columns in resultind dataframe:

In [8]:

grt = mindb.Garnet()
g.oxides.apfu(grt, keep=["Comment"])

grt = mindb.Garnet() g.oxides.apfu(grt, keep=["Comment"])

Out[8]:

	Si{4+}	Al{3+}	Mg{2+}	Fe{2+}	Mn{2+}	Ca{2+}	K{+}	Na{+}	Ti{4+}	Cr{3+}	Zn{2+}	P{5+}	Y{3+}	Fe{3+}	Comment
7	3.001661	1.981734	0.495678	2.317451	0.067297	0.122041	0.0	0.000000	0.000000	0.000000	0.0	0.0	0.000000	0.013735	g-09
8	2.989651	1.997896	0.551528	2.240917	0.067307	0.123922	0.0	0.005398	0.000778	0.002138	0.0	0.0	0.004529	0.009537	g-10

We also can quickly calculate proportions of mineral endmembers:

In [9]:

g.oxides.endmembers(grt)

g.oxides.endmembers(grt)

Out[9]:

	Alm	Prp	Sps	Grs	Adr	Uv	CaTi
7	0.771849	0.165090	0.022414	0.040367	0.000280	0.000000	0.000000
8	0.751060	0.184849	0.022558	0.041276	0.000197	0.000044	0.000016

or ignoring ferric iron and using mindb.Garnet_Fe2 model

In [10]:

grt2 = mindb.Garnet_Fe2()
g.oxides.endmembers(grt2)

grt2 = mindb.Garnet_Fe2() g.oxides.endmembers(grt2)

Out[10]:

	Alm	Prp	Sps	Grs
7	0.772888	0.164339	0.022312	0.040462
8	0.752215	0.183990	0.022454	0.041341

To recalculate iron based on charge balance, use recalculate_Fe:

In [11]:

g.oxides.recalculate_Fe(mineral=grt)

g.oxides.recalculate_Fe(mineral=grt)

Out[11]:

	SiO2	Al2O3	MgO	FeO	MnO	CaO	K2O	Na2O	TiO2	Cr2O3	ZnO	P2O5	Y2O3	Fe2O3
7	37.816	21.184	4.189	34.911088	1.001	1.435	0.0	0.000	0.000	0.000	0.0	0.006	0.000	0.229950
8	37.584	21.311	4.651	33.685819	0.999	1.454	0.0	0.035	0.013	0.034	0.0	0.030	0.107	0.232472

or you can convert iron manually:

In [12]:

g.oxides.convert_Fe(to="Fe2O3")

g.oxides.convert_Fe(to="Fe2O3")

Out[12]:

	SiO2	Al2O3	MgO	MnO	CaO	K2O	Na2O	TiO2	Cr2O3	ZnO	P2O5	Y2O3	Fe2O3
7	37.816	21.184	4.189	1.001	1.435	0.0	0.000	0.000	0.000	0.0	0.006	0.000	39.028228
8	37.584	21.311	4.651	0.999	1.454	0.0	0.035	0.013	0.034	0.0	0.030	0.107	37.669053

Similarily, we can recalculate analyses of plagioclase

In [13]:

p = df.petro.search("pl-", on="Comment")
plg = mindb.Feldspar()
p.oxides.endmembers(plg)

p = df.petro.search("pl-", on="Comment") plg = mindb.Feldspar() p.oxides.endmembers(plg)

Out[13]:

	An	Ab	Or
6	0.285439	0.712666	0.001895
9	0.287517	0.706279	0.006204
10	0.284836	0.710546	0.004619

Mineral plots¶

petroplots accessor provides some common plots, e.g. garnet profile:

In [14]:

em = pro.oxides.endmembers(grt2)
em.petroplots.profile()

em = pro.oxides.endmembers(grt2) em.petroplots.profile()

we can use available keyword arguments to improve the plot

In [15]:

em.petroplots.profile(percents=True, extra=["Alm"], lim_extra=(25, 75), markers=['o', 'd', 'v', 's'])

em.petroplots.profile(percents=True, extra=["Alm"], lim_extra=(25, 75), markers=['o', 'd', 'v', 's'])

or ternary plots:

In [16]:

em.petroplots.ternary("Alm", "Sps", "Grs", c="Prp")

em.petroplots.ternary("Alm", "Sps", "Grs", c="Prp")

In [17]:

rock[["MgO", "SiO2", "Al2O3"]].petroplots.ternary()

rock[["MgO", "SiO2", "Al2O3"]].petroplots.ternary()

with some modifications

In [18]:

rock[["MgO", "SiO2", "Al2O3"]].petroplots.ternary(tlim=(0, 0.3), llim=(0.5, 1), grid=True)

rock[["MgO", "SiO2", "Al2O3"]].petroplots.ternary(tlim=(0, 0.3), llim=(0.5, 1), grid=True)

EPMA area analyses¶

Here we use example data of local bulk rock composition measured from thinsection. To obtain average composition, several areas of the same size have been measured:

In [19]:

rock

rock

Out[19]:

	Label	Na2O	MgO	Al2O3	SiO2	P2O5	K2O	CaO	TiO2	MnO	FeO	Total
0	Spectrum 1	3.33	1.46	9.32	77.82	0.00	1.84	0.90	0.75	0.06	3.96	99.35
1	Spectrum 2	4.18	2.05	13.19	73.44	0.03	2.64	1.13	1.07	0.07	5.81	103.62
2	Spectrum 3	2.31	6.25	20.73	39.89	0.33	6.35	0.89	3.53	0.83	19.01	100.11
3	Spectrum 4	1.89	4.06	20.62	38.54	0.74	2.20	2.01	3.73	4.62	24.50	102.90
4	Spectrum 5	3.80	3.97	21.22	43.56	0.31	3.65	1.48	3.29	1.84	17.02	100.15
5	Spectrum 6	2.67	4.66	20.09	41.00	0.30	3.89	1.45	2.93	2.84	20.59	100.42
6	Spectrum 7	0.57	3.86	19.62	35.74	0.51	1.29	1.99	3.48	6.68	28.32	102.06
7	Spectrum 8	1.54	6.96	19.39	38.36	0.24	7.05	0.67	3.63	0.97	20.72	99.52
8	Spectrum 9	3.06	1.39	8.66	80.20	0.00	1.74	0.85	0.74	0.05	4.08	100.70

To calculate the average composition, we can calculate the average using mean

In [20]:

avg = rock.oxides.mean()
avg

avg = rock.oxides.mean() avg

Out[20]:

	Na2O	MgO	Al2O3	SiO2	P2O5	K2O	CaO	TiO2	MnO	FeO
0	2.594444	3.851111	16.982222	52.061111	0.273333	3.405556	1.263333	2.572222	1.995556	16.001111

To use this composition for modelling software, you can use methods to format it. Note that apatite correction and iron conversion to FeO is done automatically. For THERMOCALC

In [21]:

avg.oxides.TCbulk(H2O=1.5, oxygen=0.1)

avg.oxides.TCbulk(H2O=1.5, oxygen=0.1)

bulk    H2O   SiO2  Al2O3    CaO    MgO    FeO    K2O   Na2O   TiO2    MnO      O
bulk  5.329 54.420 10.461  1.012  6.001 13.988  2.271  2.629  2.023  1.767  0.100  % 0

for PerpleX

In [22]:

avg.oxides.Perplexbulk(H2O=1.5, oxygen=0.1)

avg.oxides.Perplexbulk(H2O=1.5, oxygen=0.1)

begin thermodynamic component list
H2O   1  5.32855      0.00000      0.00000     molar amount
SiO2  1 54.42028      0.00000      0.00000     molar amount
Al2O3 1 10.46076      0.00000      0.00000     molar amount
CaO   1  1.01178      0.00000      0.00000     molar amount
MgO   1  6.00118      0.00000      0.00000     molar amount
FeO   1 13.98809      0.00000      0.00000     molar amount
K2O   1  2.27068      0.00000      0.00000     molar amount
Na2O  1  2.62907      0.00000      0.00000     molar amount
TiO2  1  2.02279      0.00000      0.00000     molar amount
MnO   1  1.76681      0.00000      0.00000     molar amount
O2    1  0.20000      0.00000      0.00000     molar amount
end thermodynamic component list

or MAGEMin

In [23]:

avg.oxides.MAGEMin(H2O=1.5, oxygen=0.1, db='mp')

avg.oxides.MAGEMin(H2O=1.5, oxygen=0.1, db='mp')

# HEADER
title; comments; db; sysUnit; oxide; frac; frac2
# BULK-ROCK COMPOSITION
0;petropandas;mp;mol;[SiO2, Al2O3, CaO, MgO, FeO, K2O, Na2O, TiO2, O, MnO, H2O];[54.420, 10.461, 1.012, 6.001, 13.988, 2.271, 2.629, 2.023, 0.100, 1.767, 5.329];

In [ ]: