Bayesian inference for a skew-normal IRT model under the centred parameterization

Item response theory (IRT) comprises a set of statistical models which are useful in many fields, especially when there is interest in studying latent variables. These latent variables are directly considered in the Item Response Models (IRM) and they are usually called latent traits. A usual assumption for parameter estimation of the IRM, considering one group of examinees, is to assume that the latent traits are random variables which follow a standard normal distribution. However, many works suggest that this assumption does not apply in many cases. Furthermore, when this assumption does not hold, the parameter estimates tend to be biased and misleading inference can be obtained. Therefore, it is important to model the distribution of the latent traits properly. In this paper we present an alternative latent traits modeling based on the so-called skew-normal distribution; see Genton (2004). We used the centred parameterization, which was proposed by Azzalini (1985). This approach ensures the model identifiability as pointed out by Azevedo et al. (2009b). Also, a Metropolis Hastings within Gibbs sampling (MHWGS) algorithm was built for parameter estimation by using an augmented data approach. A simulation study was performed in order to assess the parameter recovery in the proposed model and the estimation method, and the effect of the asymmetry level of the latent traits distribution on the parameter estimation. Also, a comparison of our approach with other estimation methods (which consider the assumption of symmetric normality for the latent traits distribution) was considered. The results indicated that our proposed algorithm recovers properly all parameters. Specifically, the greater the asymmetry level, the better the performance of our approach compared with other approaches, mainly in the presence of small sample sizes (number of examinees). Furthermore, we analyzed a real data set which presents indication of asymmetry concerning the latent traits distribution. The results obtained by using our approach confirmed the presence of strong negative asymmetry of the latent traits distribution. (C) 2010 Elsevier B.V. All rights reserved.

RNA Isolation method for polysaccharide rich algae: agar producing Gracilaria tenuistipitata (Rhodophyta)

RNA isolation is essential to study gene expression at the molecular level. However, RNA isolation is difficult in organisms (plants and algae) that contain large amounts of polysaccharides, which co-precipitate with RNA. Currently, there is no commercial kit available, specifically for the isolation of high-quality RNA from these organisms. Furthermore, because of the large amounts of polysaccharides, the common protocols for RNA isolation usually result in poor yields when applied to algae. Here we describe a simple method for RNA isolation from the marine red macroalga Gracilaria tenuistipitata var. liui Zhang et Xia (Rhodophyta), which can be applied to other plants and algae.

Predicting and characterizing new molecular species at a high-level theoretical approach: HSeCl, HClSe, and SeCl

New molecular species HSeCl, HClSe, and SeCl were investigated at a high level of theory, CCSD(T), with a series of correlation consistent basis sets with extrapolation to the CBS limit. Account has been taken for valence-only and core-valence correlation effects, and of anharmonic effects on the vibrational frequencies. HSeCl is 43.25 kcal mol (1) more stable than HClSe. A barrier (Delta G(#)) of 47.20 kcal mol (1) separates these species. Internuclear distances are generally overestimated by 0.008 angstrom in the valence-only correlation calculations. Inclusion of anharmonicity leads to much improved vibrational frequencies. For SeCl, we estimate Delta H(f) (0 K) = 23.96 and Delta H(f) (298.15 K) = 24.64 kcal mol (1); for HSeCl, we had 4.20 and 4.97 kcal mol (1), respectively. (C) 2010 Elsevier B.V. All rights reserved.

Development of a sequential injection-square wave voltammetry method for determination of paraquat in water samples employing the hanging mercury drop electrode

This work describes the development and optimization of a sequential injection method to automate the determination of paraquat by square-wave voltammetry employing a hanging mercury drop electrode. Automation by sequential injection enhanced the sampling throughput, improving the sensitivity and precision of the measurements as a consequence of the highly reproducible and efficient conditions of mass transport of the analyte toward the electrode surface. For instance, 212 analyses can be made per hour if the sample/standard solution is prepared off-line and the sequential injection system is used just to inject the solution towards the flow cell. In-line sample conditioning reduces the sampling frequency to 44 h(-1). Experiments were performed in 0.10 M NaCl, which was the carrier solution, using a frequency of 200 Hz, a pulse height of 25 mV, a potential step of 2 mV, and a flow rate of 100 mu L s(-1). For a concentration range between 0.010 and 0.25 mg L(-1), the current (i(p), mu A) read at the potential corresponding to the peak maximum fitted the following linear equation with the paraquat concentration (mg L(-1)): ip = (-20.5 +/- 0.3) Cparaquat -(0.02 +/- 0.03). The limits of detection and quantification were 2.0 and 7.0 mu g L(-1), respectively. The accuracy of the method was evaluated by recovery studies using spiked water samples that were also analyzed by molecular absorption spectrophotometry after reduction of paraquat with sodium dithionite in an alkaline medium. No evidence of statistically significant differences between the two methods was observed at the 95% confidence level.

Calculating Flash Point Numbers from Molecular Structure: An Improved Method for Predicting the Flash Points of Acyclic Alkanes

We report a novel method for calculating flash points of acyclic alkanes from flash point numbers, N(FP), which can be calculated from experimental or calculated boiling point numbers (Y(BP)) with the equation N(FP) = 1.020Y(BP) - 1.083 Flash points (FP) are then determined from the relationship FP(K) = 23.369N(FP)(2/3) + 20.010N(FP)(1/3) + 31.901 For it data set of 102 linear and branched alkanes, the correlation of literature and predicted flash points has R(2) = 0.985 and an average absolute deviation of 3.38 K. N(FP) values can also be estimated directly from molecular structure to produce an even closer correspondence of literature and predicted FP values. Furthermore, N(FP) values provide a new method to evaluate the reliability of literature flash point data.

Isomers on the [H, S(2), Cl] potential energy surface: A high level investigation

This study reports a systematic state-of-the-art characterization of new sulfur-chlorine species on the [H, S(2), Cl] potential energy surface. Coupled cluster theory singles and doubles with perturbative contributions of connected triples, using the series of correlation consistent basis sets with extrapolations to the complete basis set limit (CBS), were employed to quantify the energetic quantities involved in the isomerization processes on this surface. The structures and vibrational frequencies are unique for some species and represent the most accurate investigation to date. These molecules are potentially a new route of coupling the sulfur and chlorine chemistries in the atmosphere, and conditions of high concentration of H(2)S (HS) like in volcanic eruptions might contribute to their formation. Also an assessment of the MP2/CBS approach relative to CCSD(T)/CBS provides insights on the expected performance of MP2/CBS on the characterization of polysulfides, and also of more complex systems containing disulfide bridges. (C) 2009 Elsevier B.V. All rights reserved.

A method for Ca, Fe, Ga, Na, Si and Zn determination in alumina by inductively coupled plasma optical emission spectrometry after aluminum precipitation

In this present work a method for the determination of Ca, Fe, Ga, Na, Si and Zn in alumina (Al(2)O(3)) by inductively coupled plasma optical emission spectrometry (ICP OES) with axial viewing is presented. Preliminary studies revealed intense aluminum spectral interference over the majority of elements and reaction between aluminum and quartz to form aluminosilicate, reducing drastically the lifetime of the torch. To overcome these problems alumina samples (250 mg) were dissolved with 5 mL HCl + 1.5 mLH(2)SO(4) + 1.5 mL H(2)O in a microwave oven. After complete dissolution the volume was completed to 20 mL and aluminum was precipitated as Al(OH)(3) with NH(3) (by bubbling NH(3) into the solution up to a pH similar to 8, for 10 min). The use of internal standards (Fe/Be, Ga/Dy, Zn/In and Na/Sc) was essential to obtain precise and accurate results. The reliability of the proposed method was checked by analysis of alumina certified reference material (Alumina Reduction Grade-699, NIST). The found concentrations (0.037%w(-1) CaO, 0.013% w w(-1) Fe(2)O(3), 0.012%w w(-1)Ga(2)O(3), 0.49% w w(-1) Na(2)O, 0.014% w w(-1) SiO(2) and 0.013% w w(-1) ZnO) presented no statistical differences compared to the certified values at a 95% confidence level. (C) 2011 Elsevier B.V. All rights reserved.

Simple Method to Evaluate and to Predict Flash Points of Organic Compounds

Flash points (T(FP)) of organic compounds are calculated from their flash point numbers, N(FP), with the relationship T(FP) = 23.369N(FP)(2/3) + 20.010N(FP)(1/3) + 31.901. In turn, the N(FP) values can be predicted from boiling point numbers (Y(BP)) and functional group counts with the equation N(FP) = 0.974Y(BP) + Sigma(i)n(i)G(i) + 0.095 where G(i) is a functional group-specific contribution to the value of N(FP) and n(i) is the number of such functional groups in the structure. For a data set consisting of 1000 diverse organic compounds, the average absolute deviation between reported and predicted flash points was less than 2.5 K.

New Molecular Species of Potential Interest to Atmospheric Chemistry: Isomers on the [H, S(2), Br] Potential Energy Surface

This work reports a state-of-the-art theoretical characterization of four new sulfur-bromine species and five transition states on the [H, S(2), Br] potential energy surface. Our highest level theoretical approach employed the method coupled cluster singles and doubles with perturbative contributions of connected triples, CCSD(T), along with the series of correlation-consistent basis sets and with extrapolation to the complete basis set (CBS) limit in the optimization of the geometrical parameters and to quantify the energetic quantities. The structural and vibrational frequencies here reported are unique and represent the most accurate investigation to date of these species. The global minimum corresponds to a skewed structure HSSBr with a disulfide bond; this is followed by a pyramidal-like structure, SSHBr, 18.85 kcal/mol above the minimum. Much higher in energy, we found another skewed structure, HSBrS (50.29 kcal/mol), with one S-Br dative-type bond, and another pyramidal-like one, HBrSS (109.80 kcal/mol), with two S-Br dative-type bonds. The interconversion of HSSBr into SSHBr can occur via a transfer of either the hydrogen or the bromine atom but involves a very high barrier of about 43 kcal/mol. These molecules are potentially a new route of coupling the sulfur and bromine chemistry in the atmosphere, and conditions of high concentration of H(2)S like in volcanic eruptions might contribute to their formation. We note that HSSBr can act as a reservoir molecule for the reaction between the radicals HSS and Br. Also, an assessment of the methods DFT/B3LYP/CBS and MP2/CBS relative to CCSD(T)/CBS provides insights on the expected performance of these methods on the characterization of polysulfides and also of more complex systems containing disulfide bridges.

A quantum chemical study on a set of non-imidazole H(3) antihistamine molecules

Molecular orbital calculations were carried out on a set of 28 non-imidazole H(3) antihistamine compounds using the Hartree-Fock method in order to investigate the possible relationships between electronic structural properties and binding affinity for H3 receptors (pK(i)). It was observed that the frontier effective-for-reaction molecular orbital (FERMO) energies were better correlated with pK(i) values than highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy values. Exploratory data analysis through hierarchical cluster (HCA) and principal component analysis (PCA) showed a separation of the compounds in two sets, one grouping the molecules with high pK(i) values, the other gathering low pK(i) value compounds. This separation was obtained with the use of the following descriptors: FERMO energies (epsilon(FERMO)), charges derived from the electrostatic potential on the nitrogen atom (N(1)), electronic density indexes for FERMO on the N(1) atom (Sigma((FERMO))c(i)(2)). and electrophilicity (omega`). These electronic descriptors were used to construct a quantitative structure-activity relationship (QSAR) model through the partial least-squares (PLS) method with three principal components. This model generated Q(2) = 0.88 and R(2) = 0.927 values obtained from a training set and external validation of 23 and 5 molecules, respectively. After the analysis of the PLS regression equation and the values for the selected electronic descriptors, it is suggested that high values of FERMO energies and of Sigma((FERMO))c(i)(2), together with low values of electrophilicity and pronounced negative charges on N(1) appear as desirable properties for the conception of new molecules which might have high binding affinity. 2010 Elsevier Inc. All rights reserved.

Classification of Parkinson’s Disease using MultiPass Lvq,Logistic Model Tree,K-Star for Audio Data set : Classification of Parkinson Disease using Audio Dataset

Parkinson's disease (PD) is a degenerative illness whose cardinal symptoms include rigidity, tremor, and slowness of movement. In addition to its widely recognized effects PD can have a profound effect on speech and voice.The speech symptoms most commonly demonstrated by patients with PD are reduced vocal loudness, monopitch, disruptions of voice quality, and abnormally fast rate of speech. This cluster of speech symptoms is often termed Hypokinetic Dysarthria.The disease can be difficult to diagnose accurately, especially in its early stages, due to this reason, automatic techniques based on Artificial Intelligence should increase the diagnosing accuracy and to help the doctors make better decisions. The aim of the thesis work is to predict the PD based on the audio files collected from various patients.Audio files are preprocessed in order to attain the features.The preprocessed data contains 23 attributes and 195 instances. On an average there are six voice recordings per person, By using data compression technique such as Discrete Cosine Transform (DCT) number of instances can be minimized, after data compression, attribute selection is done using several WEKA build in methods such as ChiSquared, GainRatio, Infogain after identifying the important attributes, we evaluate attributes one by one by using stepwise regression.Based on the selected attributes we process in WEKA by using cost sensitive classifier with various algorithms like MultiPass LVQ, Logistic Model Tree(LMT), K-Star.The classified results shows on an average 80%.By using this features 95% approximate classification of PD is acheived.This shows that using the audio dataset, PD could be predicted with a higher level of accuracy.

The impact of big-boxes on local retail : What happens when IKEA comes to town?

The development of large discount retailers, or big-boxes as they are sometimes referred to, are often subject to heated debate and their entry on a market is greeted with either great enthusiasm or dread. For instance, the world’s largest retailer Wal-Mart (Forbes 2014) has a number of anti- and pro-groups dedicated to its being and the event of a Wal-Mart entry tends to be met with protests and campaigns (Decamme 2013) but also welcomed by, for instance, consumers (Davis & DeBonis 2013). Also in Sweden, the entry of a big box is a hot topic and before IKEA’s opening i Borlänge 2013, the first in Sweden in more than five years, great expectations were mixed with worry (Västerbottens-Kuriren 2011).The presence of large scale discount retailers is not, however, a novel phenomenon but a part of a long-term change in retailing that has taken place globally over the past couple of decades (Taylor & Smalling, 2005). As noted by Dawson (2006), the trend in Europe has over the past few decades gone towards an increasing concentration of large firms along with a decrease of smaller firms.This trend is also detectable in the Swedish retail industry. Over the past decade, the retailing industry in Sweden has increased by around 190 Billion SEK, and its share of GDP has risen from 2,7% to 2,9%, while the number of employees have increased from 200 000 to 250 000 (HUI 2013). This growth, however, has not been distributed evenly but rather it has been oriented mainly towards out-of-town retail clusters. Parallel to this development, the number of large retailers has risen at the expense of market shares of smaller independent firms (Rämme et al 2010). Thereby, the presence of large scale retailers is simply part of a changing retail landscape.The effects of this development, where large scale retailing agents relocate shopping to out-of-town shopping areas, have been heavily debated. On the one hand, the big-boxes are accused of displacing independent small retail businesses in the city-centers and the residential areas, resulting in, to some extent, reduced employment opportunities and less availability for the consumers - especially the elderly (Ljungberg et al 2006). In addition, as access to shopping now tends to require some sort of a motorized vehicle, environmental aspects to the discussion have emerged. Ultimately these types of concerns have resulted in calls for regulations against this development (Olsson 2010). On the other hand, the proponents of the new shopping landscape argue that this evolution implies productivity gains, the benefits of lower prices and an increased variety of products (Maican & Orth 2012). Moreover it is argued that it leads to, for instance, better services (such as longer opening hours) and a creative destruction transformation pressure on retailers, which brings about a renewal of city-centerIIretail and services, increasing their attractivity (Bergström 2010). The belief in benefits of a big box entry can be exemplified by the attractivity of IKEA, and the fact that municipalities are prepared to commit to expenses amounting up to hundreds of millions in order to attract the entry of this big-box. Borlänge municipality, for instance, agreed to expenses of about 350 million SEK in order to secure the entry of IKEA, which opened in 2013 (Blomgren 2009).Against this backdrop, the overall effects of large discount retailers become important: Are the economic benefits enough to warrant subsidies or are there, on the contrary, some very compelling grounds for regulations against these types of establishments? In other words; how is overall retail in a region where a store like IKEA enters affected? And how are local retail firms affected?In order to answer these questions, the purpose of this thesis is to study how entry of a big-box retailer affects the entry region. The object of this study is IKEA - one of the world’s largest retailers, with 345 stores, active in over 40 countries and with profits of about 3.3 billion (IKEA 2013; IKEA 2014). By studying the effects of IKEA-entry, both on an aggregated level and on firm level, this thesis intends to find indications of how large discount retail establishments in general can be expected to affect the economic development both in a region overall, but also on the local firm level, something which is of interest to both policymakers as well as the retailing industry in general.The first paper examines the effects of IKEA on retail revenues and employment in the municipalities that IKEA chose to enter between 2000 and 2011; Gothenburg, Haparanda, Kalmar and Karlstad. By means of a matching method we first identify non-entry municipalities that have a similar probability of IKEA entry as the true entry municipalities. Then, using these non-entry municipalities as a control group, the causal effects of IKEA entry can be estimated using a treatment-control approach. We also extend the analysis to examine the spatial impact of IKEA by estimating the effects on retail in neighboring municipalities. It is found that a new IKEA store increases revenues in durable goods trade with 20% in the entry municipality and the number of employees with 17%. Only small, and in most cases statistically insignificant, negative effects were found in neighboring municipalities.It appears that there is a positive net effect on durables retail sales and employment in the entry municipality. However, the analysis is based on data on an aggregated municipality level and thereby it remains unclear if and how the effects vary within the entry municipalities. In addition, the data used in the first study includes the sales and employment of IKEA itself, which could account for the majority of the increases in employment and retail. Thereby the potential spillover effects on incumbent retailers in the entry municipalities cannot be discerned in the first study.IIITo examine effects of IKEA entry on incumbent retail firms, the second paper in this thesis analyses how IKEA entry affects the revenues and employment of local retail firms in three municipalities; Haparanda, Kalmar and Karlstad, which experienced entry by IKEA between 2000 and 2010. In this second study, we exclude Gothenburg due to the fact that big-box entry appears to have weaker effects in metropolitan areas (as indicated by Artz & Stone 2006). By excluding Gothenburg we aim to reduce the geographical heterogeneity in our study. We obtain control municipalities that are as similar as possible to the three entry municipalities using the same method as in the previous study, but including a slightly different set of variables in the selection equation. Using similar retail firms in the control municipalities as our comparison group, we estimate the impact of IKEA entry on revenues and employment for retail firms located at varying distances from the IKEA entry site.The results generated in this study imply that entry by IKEA increases revenues in incumbent retail firms by, on average, 11% in the entry municipalities. In addition, we do not find any significant impact on retail revenues in the city centers of the entry municipalities. However, we do find that retail firms within 1 km of the IKEA experience increases in revenues of about 26%, which indicates large spillover effects in the area nearby the entry site. As expected, this impact decreases as we expand the buffer zone: firms located between 0-2 km experiences a 14% increase and firms in 2-5 km experiences an increase of 10%. We do not find any significant impacts on retail employment.

Digital spiral analysis for objective assessment of fine motor timing variability in Parkinson's disease

OBJECTIVES: To develop a method for objective assessment of fine motor timing variability in Parkinson’s disease (PD) patients, using digital spiral data gathered by a touch screen device. BACKGROUND: A retrospective analysis was conducted on data from 105 subjects including65 patients with advanced PD (group A), 15 intermediate patients experiencing motor fluctuations (group I), 15 early stage patients (group S), and 10 healthy elderly subjects (HE) were examined. The subjects were asked to perform repeated upper limb motor tasks by tracing a pre-drawn Archimedes spiral as shown on the screen of the device. The spiral tracing test was performed using an ergonomic pen stylus, using dominant hand. The test was repeated three times per test occasion and the subjects were instructed to complete it within 10 seconds. Digital spiral data including stylus position (x-ycoordinates) and timestamps (milliseconds) were collected and used in subsequent analysis. The total number of observations with the test battery were as follows: Swedish group (n=10079), Italian I group (n=822), Italian S group (n = 811), and HE (n=299). METHODS: The raw spiral data were processed with three data processing methods. To quantify motor timing variability during spiral drawing tasks Approximate Entropy (APEN) method was applied on digitized spiral data. APEN is designed to capture the amount of irregularity or complexity in time series. APEN requires determination of two parameters, namely, the window size and similarity measure. In our work and after experimentation, window size was set to 4 and similarity measure to 0.2 (20% of the standard deviation of the time series). The final score obtained by APEN was normalized by total drawing completion time and used in subsequent analysis. The score generated by this method is hence on denoted APEN. In addition, two more methods were applied on digital spiral data and their scores were used in subsequent analysis. The first method was based on Digital Wavelet Transform and Principal Component Analysis and generated a score representing spiral drawing impairment. The score generated by this method is hence on denoted WAV. The second method was based on standard deviation of frequency filtered drawing velocity. The score generated by this method is hence on denoted SDDV. Linear mixed-effects (LME) models were used to evaluate mean differences of the spiral scores of the three methods across the four subject groups. Test-retest reliability of the three scores was assessed after taking mean of the three possible correlations (Spearman’s rank coefficients) between the three test trials. Internal consistency of the methods was assessed by calculating correlations between their scores. RESULTS: When comparing mean spiral scores between the four subject groups, the APEN scores were different between HE subjects and three patient groups (P=0.626 for S group with 9.9% mean value difference, P=0.089 for I group with 30.2%, and P=0.0019 for A group with 44.1%). However, there were no significant differences in mean scores of the other two methods, except for the WAV between the HE and A groups (P<0.001). WAV and SDDV were highly and significantly correlated to each other with a coefficient of 0.69. However, APEN was not correlated to neither WAV nor SDDV with coefficients of 0.11 and 0.12, respectively. Test-retest reliability coefficients of the three scores were as follows: APEN (0.9), WAV(0.83) and SD-DV (0.55). CONCLUSIONS: The results show that the digital spiral analysis-based objective APEN measure is able to significantly differentiate the healthy subjects from patients at advanced level. In contrast to the other two methods (WAV and SDDV) that are designed to quantify dyskinesias (over-medications), this method can be useful for characterizing Off symptoms in PD. The APEN was not correlated to none of the other two methods indicating that it measures a different construct of upper limb motor function in PD patients than WAV and SDDV. The APEN also had a better test-retest reliability indicating that it is more stable and consistent over time than WAV and SDDV.

Genetic heterogeneity of residual variance : estimation of variance components using double hierarchical generalized linear models

Background: The sensitivity to microenvironmental changes varies among animals and may be under genetic control. It is essential to take this element into account when aiming at breeding robust farm animals. Here, linear mixed models with genetic effects in the residual variance part of the model can be used. Such models have previously been fitted using EM and MCMC algorithms. Results: We propose the use of double hierarchical generalized linear models (DHGLM), where the squared residuals are assumed to be gamma distributed and the residual variance is fitted using a generalized linear model. The algorithm iterates between two sets of mixed model equations, one on the level of observations and one on the level of variances. The method was validated using simulations and also by re-analyzing a data set on pig litter size that was previously analyzed using a Bayesian approach. The pig litter size data contained 10,060 records from 4,149 sows. The DHGLM was implemented using the ASReml software and the algorithm converged within three minutes on a Linux server. The estimates were similar to those previously obtained using Bayesian methodology, especially the variance components in the residual variance part of the model. Conclusions: We have shown that variance components in the residual variance part of a linear mixed model can be estimated using a DHGLM approach. The method enables analyses of animal models with large numbers of observations. An important future development of the DHGLM methodology is to include the genetic correlation between the random effects in the mean and residual variance parts of the model as a parameter of the DHGLM.

L1 Use in the EFL-classroom : A Literature Review on Teachers’ Use of the L1 in the EFL-classroom on Upper Secondary Level

Even though English should mainly be used in the EFL-classroom according to the Swedish national curriculum, some recent scholars have argued that a judicious use of the students’ L1 by the teachers in some particular situations may benefit the learning environment there. From this context, this thesis examines what research says about in what particular situations the L1 is used by teachers in upper secondary EFL-classrooms as well as the teachers’ and the students’ attitudes towards this practice. The method used was a systematic literature review, where seven articles from all across the globe were analyzed, compared and synthesized. The results show that the L1 was mainly used by the teachers when managing discipline, when explaining grammar and when teaching vocabulary. However, the articles did not conclude how the L1 could be used strategically by the teachers. Concerning the teachers’ and students’ attitudes, the majority of these were positive towards L1 use by the teacher in the above mentioned situations. However, the teachers were not aware of how the L1 could be used by them in a judicious and a strategic way. Lastly, it can be concluded that more research is needed on how the L1 can be used more strategically by the teachers as well as on the students’ perspective on this.