6. Plotting

Generalistic plots for data analysis.

6.1. Boxplots

A variety of boxplots for analysis of point data.

6.1.1. Boxplots

resourcegeo.plots.boxplots.data_by_category(data, var, catcol, tmin=None, use_all=True)

Convert df with category to list of arrays vals + labels.

Return: df_values (list[np.array]): array-values for each category df_labels (list[str]) : Categorical labels

resourcegeo.plots.boxplots.boxplot_by_category(data, var, catcol, cats=None, tmin=None, ylabel=None, xlabel=None, title=None, flname=None, markersize_flier=2, markersize_mean=4, widths=None, figsize=None, ylim=None, stats=True, stats_ycoord=None, use_all=True, yscale='linear')

Univariate. Boxplots by category. Optionally add number and unweighted mean to the plot.

Parameters:

  • data (pd.DataFrame) – df with values and a category column

  • var (str) – column name for variable

  • catcol (str) – column name for category column

  • cats (optional) – Categories to consider from catcol

  • xlabel (str, optional) – x-label

  • ylabel (str, optional) – y-label

  • figsize (tuple,optional) – figsize

  • flname (str, optional) – path to save file if not None

  • stats (bool,optional) – Show number and unweighted mean for each box plot

  • yscale (str, optional) – y-axis scale

  • stats_ycoord (float, optional) – bottom coordinate to add the stats (count and mean). The mean is unweighted.

  • tmin (float, optional) – Minumum var value to filter out values

  • use_all (bool, optional) – If False, plot boxplots only for categories with >0 number of data. If True, it shows all boxplots. If tmin is not None, then tmin is applied before use_all.

Examples:

import resourcegeo as rs
df = rs.BaseData('assay_geo').data
_ = rs.boxplot_by_category(data = df,var = 'CUpc',catcol = 'UNIT',
        tmin=0, markersize_flier=0.5,widths=0.7,markersize_mean=4,
        ylim=(-0.2,2), stats_ycoord=-0.15, figsize=(17,6),stats=False, use_all=False)
_images/plotting-1.png
resourcegeo.plots.boxplots.boxplot_by_category_sensitivity(dfs, var, catcol, yscale='linear', markersize_flier=2, markersize_mean=4, flname=None, figsize=None, title=None, xlabel=None, ylabel=None, bulkstats=True, coverage=0.75, stats_ycoord=0, ylim=None)

Univariate. Grouped boxplots by categories. It can be used to analyze sensitivity results.

Parameters:

  • dfs (list[pd.DataFrame]) – list of DataFrame(s)

  • var (str) – shared variable column-name in all df in dfs

  • catcol (str) – shared category column in all df in dfs

  • title (str, optional) – title of the graph

  • xlabel (str, optional) – x-label

  • ylabel (str, optional) – y-label

  • flname (str., optional) – path to save file if not None

  • figsize (tuple, optional) – figsize

  • bulkstats (bool, optional) – show number and unweighted mean

  • ylim (tuple, optional) – y-axis coordinate limits. If bulkstats is True, you may need to use a lower y-minimum coordinate.

  • stats_ycoord (float) – bottom coordinate to add the stats (count and mean). The mean is unweighted.

  • coverage (float, optional) – width of boxes expresed as fraction from 0 to 1.

Examples:

import resourcegeo as rs
df = rs.BaseData('assay_geo').data
df2 = df.loc[df['UNIT'].isin([f'{i}' for i in range(1,9)]  + ['OVB'])].copy()
df3 = df2.copy()
df3['CUpc'] = df3['CUpc'] +0.1
df4 = df2.copy()
df4['CUpc'] = df3['CUpc'] +0.2
dfs = [df2,df3,df4]
rs.boxplot_by_category_sensitivity(dfs,'CUpc','UNIT',figsize=(14,4),
        coverage=0.9,ylim=(-0.7,2), bulkstats=True,stats_ycoord=-0.6)
_images/plotting-2.png

6.2. Bar Charts

Bar Charts.

6.2.1. Drilling

resourcegeo.plots.drilling.drilling_by_year(df, var, cat, cat2=None, xlabel=None, ylabel=None, title=None, flname=None, figsize=(14, 4), cat2_colors=None, legend_loc=None, width=1)
Generates a stacked bar plot by based on a continuous variable

and two categorical columns.

Parameters:

  • df (pd.DataFrame) – df with continuous variable and two categorical columns

  • var (float) – Continuous variable to summarize sums

  • cat (str) – Column name for categorical that contains a sequence of integers to use in the bar chart. e.g. years

  • cat2 (str) – Column name for categorical values to be stacked

  • xlabel (str,optional) – x-axis label

  • ylabel (str,optional) – y-axis label

  • title (str,optional) – title for the plot

  • flname (str,optional) – path to save image

  • figsize (tuple) – figure size

  • cat2_colors (dict) – dictionary for user defined colors.

  • legend_loc (int) – position of the legend

  • width (float) – with of the bars in the bar chart

Examples:

import resourcegeo as rs
df = rs.BaseData('collar').data
drill_type= 'drilltype'
df.loc[df[drill_type].isna(),drill_type]= 'Undefined'
rs.drilling_by_year(df,var='length',cat='year',cat2=drill_type,
   xlabel='Collar year',ylabel='Length Drilled')
_images/plotting-3.png

6.3. Scatters

Scatter plots.

6.3.1. Resource Model

resourcegeo.plots.tongrad.tonnage_grade_curve(dfs, var_col, cutoffs=None, weight_col=None, as_fraction=True, ylim1=None, ylim2=None, xlim=None, title=None, flname=None, xlabel=None, ylabel=None, ylabel2=None, fontsize=11, lw=1, figsize=(5, 4))

Generate tonnage-grade curve for one or multiple dfs, for multiple variables within each df.

Parameters:

  • dfs (df or list[df]) – Data with a at least one column of values

  • var_col (str or list[list[str]]) – column-name(s) of variables in a df

  • cutoffs (list[float],optional) – cut-offs values. If None, 100 cutoffs between min-max values are chosen

  • weight_col (str,optional) – column-name if all dfs with values to weight the grade values. WARINING: All df’s must have the same weight_col name.

  • as_fraction (bool, optional) – If True, it shows tonnage axis as 0-1 fraction. If False, shows the sum of the weights.

  • ylim1 (tuple) – limit values for main y-axis (weight sum)

  • ylim2 (tuple) – limit values for secondary y_axis (average grades above cutoff)

  • xlim (tuple) – limit values for x-axis (cutoffs)

  • title (str) – plot title

  • flanme (str) – path to store plot. It does not create folders

  • fontsize (float) – overall font size

  • lw (float) – line width for GT curves

  • figsize (tuple) – figsize

Returns:

None

Examples:

import resourcegeo as rs
bm = rs.BaseData('bm').data
rs.tonnage_grade_curve(bm, var_col='grade1')
_images/plotting-4.png

6.3.2. Distributions

resourcegeo.plots.qqplot.qq_plot(d1, d2, percentiles=None, ms=1, xlim=None, ylim=None, flname=None, ylabel=None, title=None, xlabel=None, figsize=None, fontdict=None)

QQ-plot with two given distributions. Lengths of the distributions may be different. Alternatively add percentiles and summary stats.

Parameters:

  • d1 (np.array or list[float]) – distribution in x-axis

  • d2 (np.array or list[float]) – distribution in y-axis

  • percentiles (list[float]) – percentiles from 0-100 included

  • ms (float) – marker size

  • xlim (tuple[float]) – x-axis limits

  • ylim (tuple[float]) – y-axis limits

  • flname (str) – path to save plot

  • figsize (tuple(float)) – figure size

Examples:

import resourcegeo as rs
import numpy as np
rs.qq_plot(np.random.randn(500),np.random.randn(700))
_images/plotting-5.png
resourcegeo.plots.probability_plot.probability_plot(data, var, weights=None, figsize=None, xscale='log', fontsize=10, s=1, ax=None, label=None, color=None, xlim=None, title=None, xlabel=None, ylabel=None)

Probability plot of a univariate distribution.

Parameters:

  • data (pd.DataFrame) – DataFrame with values

  • var (str) – variable column-name in data

  • weights (str,optional) – column-name for weights to use

  • figsize (tuple) – Figure size

  • xscale (str, optional) – x-axis scale, it can be ‘log’ or ‘linear’

  • title (str,optional) – title

  • ylabel (str,optional) – y-axis label

  • xlabel (str,optional) – x-axis label

Examples:

import resourcegeo as rs
df = rs.BaseData('assay_geo').data
fig ,ax = plt.subplots(1,1)
_=rs.probability_plot(df,var='CUpc',ax=ax,label='Unweighted')
_=rs.probability_plot(df,var='CUpc',weights='ai',ax=ax,label='Weighted')
_images/plotting-6.png

6.4. Sensitivity

Sensitivity plots.

resourcegeo.plots.deviation.deviation_plot(x, y, base, flname=None, title=None, y1_label=None, y2_label=None, xlabel=None, hline_label='base', dev_ylim=0.2)

Plot deviations with respect a base value after performing a sensitivity analysis.

Parameters:

  • x (list[float]) – x values of the sensitivities

  • y (list[float]) – results from the multiple x-values

  • base (float) – Reference value around where deviation is calculated.

  • dev_ylim (float) – fraction of the base to use as +/- in (right) y-axis

  • xlabel (str) – x-axis label

  • y1_label (str) – Left y-axis of sensitivity results

  • y2_label (str) – Right y-axis for percentages of deviation from base value

  • title (str,optional) – title of the graph

  • flname (str,optional) – path to save figure

Returns:

None

Examples:

import resourcegeo as rs
import random
import numpy as np
n, ref = 15, 30
a = np.linspace(1,n+1,n)
s = np.random.normal(0, 0.5, n)
b = [ref+j for j in s]
rs.plots.deviation.deviation_plot(a,b,ref)
_images/plotting-7.png