Graphics Reference
In-Depth Information
Table
.
.
Correlations between the four ratio variables in Fig.
.
Cash.TA Inv.TA
CA.TA
Kap.TA
Cash.TA
.
Inv.TA
-
.
.
CA.TA
.
.
.
Kap.TA
-
.
-
.
-
.
.
gular shapes to the lower let show that the sum of the corresponding ratios is less
than a particular limit;
here. (he triangular shapes to the lower right are obtained
whenthe y variable isalwayslowerinvaluethanthe x variable, suchaswhenthecash
orinventory ratio isplotted against the current assets ratio.) henegative cash values
that have already been discussed are easily seen to the let of the bulk of the data in
the plots of cash and inventory ratios. In the plot of the current assets and property
ratios there are a few companies that surprisingly lie above the bounding diagonal -
more outlying cases to investigate. he correlation coe
cients for the variables are
given inTable
.
.Someofthemarequitehigh,andtheyarecertainly allsignificantly
different from zero, but they hint little at the structure shown. he highest value in
absolute terms is the correlation between current assets and property (-
.
). his
deserves further investigation, and Fig.
.
shows the same scatterplots as shown in
Fig.
.
,butthistimeusingthesmallestpointsizeandsome α-blendinginsteadofthe
defaults, providing a rough bivariate density estimate. It is now possible to see that,
for many of the companies, the sum of current assets and property assets is almost
equal to the Total Assets.
he scatterplots of the corresponding raw data variables provide another view of
theassociationsbetweenvariablesinthedataset,andmorestructurecanbeseenthan
with scatterplots oftheratios. Afewoftherawscatterplots exhibit afunnelstructure,
such as the plot of inventories against fixed assets and that of cash against current as-
Figure
.
.
he same scatterplots as shown in Fig.
.
, but with a smaller pointsize and α-blending to
better display the bivariate structures
